AnomalyDetection.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "qFjTq-zvR_KY"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "import numpy as np"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf"
      ],
      "metadata": {
        "id": "uMaXvpm0-q2o"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MbZHFhqt0A1U",
        "outputId": "6ba3c641-8b9a-4e1e-f5c7-0741b3fd5aaa"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "##Preliminay operations"
      ],
      "metadata": {
        "id": "sK98Slc5CcgZ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train = pd.read_csv('drive/MyDrive/Tesi/Experiment/Data/train_pods_log.csv', sep=',')"
      ],
      "metadata": {
        "id": "PnYivpJnpYib"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train.drop(labels='Unnamed: 0', axis=1, inplace=True)"
      ],
      "metadata": {
        "id": "wOICjpz8pqVY"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train.columns"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gbuhEOaUpxVh",
        "outputId": "9f407296-74c9-470a-fb8b-f7d8cdf2e5a8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['idle_node1_delta', 'iowait_node1_delta', 'user_node1_delta',\n",
              "       'system_node1_delta', 'irq_node1_delta', 'softirq_node1_delta',\n",
              "       'nice_node1_delta', 'steal_node1_delta', 'idle_node2_delta',\n",
              "       'iowait_node2_delta', 'user_node2_delta', 'system_node2_delta',\n",
              "       'irq_node2_delta', 'softirq_node2_delta', 'nice_node2_delta',\n",
              "       'steal_node2_delta', 'idle_node3_delta', 'iowait_node3_delta',\n",
              "       'user_node3_delta', 'system_node3_delta', 'irq_node3_delta',\n",
              "       'softirq_node3_delta', 'nice_node3_delta', 'steal_node3_delta',\n",
              "       'worker1_pods_mem', 'worker2_pods_mem', 'worker3_pods_mem',\n",
              "       'worker1_endpoints_mem', 'worker1_nodexp_mem', 'worker2_endpoints_mem',\n",
              "       'worker2_nodexp_mem', 'worker3_endpoints_mem', 'worker3_nodexp_mem',\n",
              "       'io_time_sdb_node1_delta', 'io_time_sdb_node2_delta',\n",
              "       'io_time_sdb_node3_delta', 'read_bytes_tot_node1_delta',\n",
              "       'read_bytes_tot_node2_delta', 'read_bytes_tot_node3_delta',\n",
              "       'reads_completed_node1_delta', 'reads_completed_node2_delta',\n",
              "       'reads_completed_node3_delta', 'worker1_ens192_delta',\n",
              "       'worker2_ens192_delta', 'worker3_ens192_delta', 'produced_logs',\n",
              "       'ts-auth-service', 'ts-travel-service', 'ts-ui-dashboard',\n",
              "       'ts-contacts-service'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 522
        },
        "id": "pMcFJA3KpmD8",
        "outputId": "95ec7f6f-7fd4-4827-c2ff-b0deaecb693e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      idle_node1_delta  iowait_node1_delta  user_node1_delta  \\\n",
              "0               445.28                0.48             19.61   \n",
              "1               444.59                0.42             21.27   \n",
              "2               446.94                0.37             19.39   \n",
              "3               443.49                0.43             21.78   \n",
              "4               444.83                0.42             20.79   \n",
              "...                ...                 ...               ...   \n",
              "5664            446.62                0.43             19.23   \n",
              "5665            445.36                0.42             20.58   \n",
              "5666            445.89                0.39             19.18   \n",
              "5667            445.62                0.40             19.67   \n",
              "5668            444.86                0.39             21.32   \n",
              "\n",
              "      system_node1_delta  irq_node1_delta  softirq_node1_delta  \\\n",
              "0                   7.79              0.0                 0.89   \n",
              "1                   7.29              0.0                 0.75   \n",
              "2                   7.19              0.0                 0.86   \n",
              "3                   7.26              0.0                 0.90   \n",
              "4                   7.43              0.0                 0.75   \n",
              "...                  ...              ...                  ...   \n",
              "5664                7.04              0.0                 0.72   \n",
              "5665                7.26              0.0                 0.80   \n",
              "5666                7.79              0.0                 0.78   \n",
              "5667                7.64              0.0                 0.77   \n",
              "5668                6.93              0.0                 0.78   \n",
              "\n",
              "      nice_node1_delta  steal_node1_delta  idle_node2_delta  \\\n",
              "0                  0.0                0.0            453.44   \n",
              "1                  0.0                0.0            454.81   \n",
              "2                  0.0                0.0            455.46   \n",
              "3                  0.0                0.0            454.75   \n",
              "4                  0.0                0.0            454.49   \n",
              "...                ...                ...               ...   \n",
              "5664               0.0                0.0            415.90   \n",
              "5665               0.0                0.0            442.33   \n",
              "5666               0.0                0.0            453.32   \n",
              "5667               0.0                0.0            452.18   \n",
              "5668               0.0                0.0            453.64   \n",
              "\n",
              "      iowait_node2_delta  ...  reads_completed_node2_delta  \\\n",
              "0                   0.47  ...                          0.0   \n",
              "1                   0.45  ...                          0.0   \n",
              "2                   0.42  ...                          0.0   \n",
              "3                   0.42  ...                          0.0   \n",
              "4                   0.42  ...                          0.0   \n",
              "...                  ...  ...                          ...   \n",
              "5664               14.62  ...                       3787.0   \n",
              "5665                3.84  ...                       1415.0   \n",
              "5666                0.46  ...                          0.0   \n",
              "5667                0.60  ...                          0.0   \n",
              "5668                0.47  ...                          5.0   \n",
              "\n",
              "      reads_completed_node3_delta  worker1_ens192_delta  worker2_ens192_delta  \\\n",
              "0                             0.0             4564712.0             5016609.0   \n",
              "1                             0.0             4464413.0             5479738.0   \n",
              "2                             0.0             4321960.0             4801737.0   \n",
              "3                             0.0             4222039.0             5719404.0   \n",
              "4                             0.0             4386564.0             5276439.0   \n",
              "...                           ...                   ...                   ...   \n",
              "5664                          0.0             4402832.0             4986912.0   \n",
              "5665                          0.0             4418401.0             5474668.0   \n",
              "5666                          0.0             4120944.0             4912318.0   \n",
              "5667                          0.0             4294250.0             5058737.0   \n",
              "5668                          0.0             4231869.0             5493439.0   \n",
              "\n",
              "      worker3_ens192_delta  produced_logs  ts-auth-service  ts-travel-service  \\\n",
              "0                9564159.0           1467              426                 20   \n",
              "1                7602923.0           1756              342                 13   \n",
              "2                7319263.0           1521              399                 23   \n",
              "3                7290660.0           1840              335                 18   \n",
              "4                7959451.0           1694              408                 21   \n",
              "...                    ...            ...              ...                ...   \n",
              "5664             7087061.0           1536              360                 19   \n",
              "5665             8096267.0           1649              341                 18   \n",
              "5666             6603181.0           1624              352                 23   \n",
              "5667             7234592.0           1594              345                 13   \n",
              "5668             8547428.0           1888              365                 20   \n",
              "\n",
              "      ts-ui-dashboard  ts-contacts-service  \n",
              "0                 249                  0.0  \n",
              "1                 201                  0.0  \n",
              "2                 234                  0.0  \n",
              "3                 196                  0.0  \n",
              "4                 239                  0.0  \n",
              "...               ...                  ...  \n",
              "5664              210                  0.0  \n",
              "5665              199                  0.0  \n",
              "5666              207                  0.0  \n",
              "5667              200                  0.0  \n",
              "5668              217                  0.0  \n",
              "\n",
              "[5669 rows x 50 columns]"
            ],
            "text/html": [
              "\n",
              "\n",
              "  <div id=\"df-d415c55c-c0cc-4076-ad5b-c1c025f336b5\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>idle_node1_delta</th>\n",
              "      <th>iowait_node1_delta</th>\n",
              "      <th>user_node1_delta</th>\n",
              "      <th>system_node1_delta</th>\n",
              "      <th>irq_node1_delta</th>\n",
              "      <th>softirq_node1_delta</th>\n",
              "      <th>nice_node1_delta</th>\n",
              "      <th>steal_node1_delta</th>\n",
              "      <th>idle_node2_delta</th>\n",
              "      <th>iowait_node2_delta</th>\n",
              "      <th>...</th>\n",
              "      <th>reads_completed_node2_delta</th>\n",
              "      <th>reads_completed_node3_delta</th>\n",
              "      <th>worker1_ens192_delta</th>\n",
              "      <th>worker2_ens192_delta</th>\n",
              "      <th>worker3_ens192_delta</th>\n",
              "      <th>produced_logs</th>\n",
              "      <th>ts-auth-service</th>\n",
              "      <th>ts-travel-service</th>\n",
              "      <th>ts-ui-dashboard</th>\n",
              "      <th>ts-contacts-service</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>445.28</td>\n",
              "      <td>0.48</td>\n",
              "      <td>19.61</td>\n",
              "      <td>7.79</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.89</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>453.44</td>\n",
              "      <td>0.47</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4564712.0</td>\n",
              "      <td>5016609.0</td>\n",
              "      <td>9564159.0</td>\n",
              "      <td>1467</td>\n",
              "      <td>426</td>\n",
              "      <td>20</td>\n",
              "      <td>249</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>444.59</td>\n",
              "      <td>0.42</td>\n",
              "      <td>21.27</td>\n",
              "      <td>7.29</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.75</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>454.81</td>\n",
              "      <td>0.45</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4464413.0</td>\n",
              "      <td>5479738.0</td>\n",
              "      <td>7602923.0</td>\n",
              "      <td>1756</td>\n",
              "      <td>342</td>\n",
              "      <td>13</td>\n",
              "      <td>201</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>446.94</td>\n",
              "      <td>0.37</td>\n",
              "      <td>19.39</td>\n",
              "      <td>7.19</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.86</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>455.46</td>\n",
              "      <td>0.42</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4321960.0</td>\n",
              "      <td>4801737.0</td>\n",
              "      <td>7319263.0</td>\n",
              "      <td>1521</td>\n",
              "      <td>399</td>\n",
              "      <td>23</td>\n",
              "      <td>234</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>443.49</td>\n",
              "      <td>0.43</td>\n",
              "      <td>21.78</td>\n",
              "      <td>7.26</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.90</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>454.75</td>\n",
              "      <td>0.42</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4222039.0</td>\n",
              "      <td>5719404.0</td>\n",
              "      <td>7290660.0</td>\n",
              "      <td>1840</td>\n",
              "      <td>335</td>\n",
              "      <td>18</td>\n",
              "      <td>196</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>444.83</td>\n",
              "      <td>0.42</td>\n",
              "      <td>20.79</td>\n",
              "      <td>7.43</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.75</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>454.49</td>\n",
              "      <td>0.42</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4386564.0</td>\n",
              "      <td>5276439.0</td>\n",
              "      <td>7959451.0</td>\n",
              "      <td>1694</td>\n",
              "      <td>408</td>\n",
              "      <td>21</td>\n",
              "      <td>239</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5664</th>\n",
              "      <td>446.62</td>\n",
              "      <td>0.43</td>\n",
              "      <td>19.23</td>\n",
              "      <td>7.04</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.72</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>415.90</td>\n",
              "      <td>14.62</td>\n",
              "      <td>...</td>\n",
              "      <td>3787.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4402832.0</td>\n",
              "      <td>4986912.0</td>\n",
              "      <td>7087061.0</td>\n",
              "      <td>1536</td>\n",
              "      <td>360</td>\n",
              "      <td>19</td>\n",
              "      <td>210</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5665</th>\n",
              "      <td>445.36</td>\n",
              "      <td>0.42</td>\n",
              "      <td>20.58</td>\n",
              "      <td>7.26</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.80</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>442.33</td>\n",
              "      <td>3.84</td>\n",
              "      <td>...</td>\n",
              "      <td>1415.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4418401.0</td>\n",
              "      <td>5474668.0</td>\n",
              "      <td>8096267.0</td>\n",
              "      <td>1649</td>\n",
              "      <td>341</td>\n",
              "      <td>18</td>\n",
              "      <td>199</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5666</th>\n",
              "      <td>445.89</td>\n",
              "      <td>0.39</td>\n",
              "      <td>19.18</td>\n",
              "      <td>7.79</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.78</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>453.32</td>\n",
              "      <td>0.46</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4120944.0</td>\n",
              "      <td>4912318.0</td>\n",
              "      <td>6603181.0</td>\n",
              "      <td>1624</td>\n",
              "      <td>352</td>\n",
              "      <td>23</td>\n",
              "      <td>207</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5667</th>\n",
              "      <td>445.62</td>\n",
              "      <td>0.40</td>\n",
              "      <td>19.67</td>\n",
              "      <td>7.64</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.77</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>452.18</td>\n",
              "      <td>0.60</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4294250.0</td>\n",
              "      <td>5058737.0</td>\n",
              "      <td>7234592.0</td>\n",
              "      <td>1594</td>\n",
              "      <td>345</td>\n",
              "      <td>13</td>\n",
              "      <td>200</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5668</th>\n",
              "      <td>444.86</td>\n",
              "      <td>0.39</td>\n",
              "      <td>21.32</td>\n",
              "      <td>6.93</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.78</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>453.64</td>\n",
              "      <td>0.47</td>\n",
              "      <td>...</td>\n",
              "      <td>5.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4231869.0</td>\n",
              "      <td>5493439.0</td>\n",
              "      <td>8547428.0</td>\n",
              "      <td>1888</td>\n",
              "      <td>365</td>\n",
              "      <td>20</td>\n",
              "      <td>217</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5669 rows × 50 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d415c55c-c0cc-4076-ad5b-c1c025f336b5')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "\n",
              "\n",
              "\n",
              "    <div id=\"df-b456a88f-ad66-4260-9329-4ecc28cddc22\">\n",
              "      <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-b456a88f-ad66-4260-9329-4ecc28cddc22')\"\n",
              "              title=\"Suggest charts.\"\n",
              "              style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "      </button>\n",
              "    </div>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "    background-color: #E8F0FE;\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: #1967D2;\n",
              "    height: 32px;\n",
              "    padding: 0 0 0 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: #E2EBFA;\n",
              "    box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: #174EA6;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "    background-color: #3B4455;\n",
              "    fill: #D2E3FC;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart:hover {\n",
              "    background-color: #434B5C;\n",
              "    box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "    filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "    fill: #FFFFFF;\n",
              "  }\n",
              "</style>\n",
              "\n",
              "    <script>\n",
              "      async function quickchart(key) {\n",
              "        const containerElement = document.querySelector('#' + key);\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      }\n",
              "    </script>\n",
              "\n",
              "      <script>\n",
              "\n",
              "function displayQuickchartButton(domScope) {\n",
              "  let quickchartButtonEl =\n",
              "    domScope.querySelector('#df-b456a88f-ad66-4260-9329-4ecc28cddc22 button.colab-df-quickchart');\n",
              "  quickchartButtonEl.style.display =\n",
              "    google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "}\n",
              "\n",
              "        displayQuickchartButton(document);\n",
              "      </script>\n",
              "      <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-d415c55c-c0cc-4076-ad5b-c1c025f336b5 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-d415c55c-c0cc-4076-ad5b-c1c025f336b5');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from keras.models import Model\n",
        "from keras.layers import Input, Dense\n",
        "\n",
        "# Load and preprocess the data\n",
        "#train.drop(labels='Time', axis=1, inplace=True)\n",
        "# Split the data into train and test sets\n",
        "train_data, val_data = train_test_split(train, test_size=0.2, random_state=12)\n",
        "val_data1, val_data2 = train_test_split(val_data, test_size=0.5, random_state=12)\n",
        "# Normalize the data using Min-Max scaling\n",
        "scaler = StandardScaler()\n",
        "train_data_scaled = scaler.fit_transform(train_data)\n",
        "val_data_scaled = scaler.fit_transform(val_data)\n",
        "val_data1_scaled = scaler.fit_transform(val_data1)\n",
        "val_data2_scaled = scaler.fit_transform(val_data2)"
      ],
      "metadata": {
        "id": "QmTDBkasVtfV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test = pd.read_csv('drive/MyDrive/Tesi/Experiment/Data/Test/Refined/test_final.csv', sep=',')\n",
        "test_pods =  pd.read_csv('drive/MyDrive/Tesi/Experiment/Data/Elastic/elastic_agg_sep_test.csv')\n",
        "test.drop(labels='Unnamed: 0', axis=1, inplace=True)\n",
        "test.drop(labels='Time', axis=1, inplace=True)\n",
        "#test.drop(labels='produced_logs', axis=1, inplace=True)\n"
      ],
      "metadata": {
        "id": "g_qurVBWFHBt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 488
        },
        "id": "2F8iOZUlEHkg",
        "outputId": "fd2bf347-5d3a-4019-e7b7-00664f887246"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     idle_node1_delta  iowait_node1_delta  user_node1_delta  \\\n",
              "0              440.24                0.45             23.28   \n",
              "1              432.90                0.50             27.34   \n",
              "2              420.56                0.67             35.13   \n",
              "3              423.56                1.43             31.65   \n",
              "4              424.19                0.52             32.19   \n",
              "..                ...                 ...               ...   \n",
              "115            443.26                0.45             22.57   \n",
              "116            446.77                0.41             19.32   \n",
              "117            419.64               24.85             21.82   \n",
              "118            445.10                0.40             20.26   \n",
              "119            441.12                0.50             24.16   \n",
              "\n",
              "     system_node1_delta  irq_node1_delta  softirq_node1_delta  \\\n",
              "0                  8.26              0.0                 1.02   \n",
              "1                 10.23              0.0                 1.98   \n",
              "2                 12.70              0.0                 3.60   \n",
              "3                 12.28              0.0                 3.77   \n",
              "4                 11.91              0.0                 3.64   \n",
              "..                  ...              ...                  ...   \n",
              "115                6.99              0.0                 0.89   \n",
              "116                7.41              0.0                 0.75   \n",
              "117                7.12              0.0                 0.85   \n",
              "118                7.48              0.0                 0.74   \n",
              "119                7.78              0.0                 0.87   \n",
              "\n",
              "     nice_node1_delta  steal_node1_delta  idle_node2_delta  \\\n",
              "0                 0.0                0.0            440.24   \n",
              "1                 0.0                0.0            432.90   \n",
              "2                 0.0                0.0            420.56   \n",
              "3                 0.0                0.0            423.56   \n",
              "4                 0.0                0.0            424.19   \n",
              "..                ...                ...               ...   \n",
              "115               0.0                0.0            443.26   \n",
              "116               0.0                0.0            446.77   \n",
              "117               0.0                0.0            419.64   \n",
              "118               0.0                0.0            445.10   \n",
              "119               0.0                0.0            441.12   \n",
              "\n",
              "     iowait_node2_delta  ...  read_bytes_tot_node1_delta  \\\n",
              "0                  0.45  ...                         0.0   \n",
              "1                  0.50  ...                         0.0   \n",
              "2                  0.67  ...                         0.0   \n",
              "3                  1.43  ...                         0.0   \n",
              "4                  0.52  ...                         0.0   \n",
              "..                  ...  ...                         ...   \n",
              "115                0.45  ...                         0.0   \n",
              "116                0.41  ...                         0.0   \n",
              "117               24.85  ...                         0.0   \n",
              "118                0.40  ...                         0.0   \n",
              "119                0.50  ...                         0.0   \n",
              "\n",
              "     read_bytes_tot_node2_delta  read_bytes_tot_node3_delta  \\\n",
              "0                           0.0                     12288.0   \n",
              "1                        4096.0                         0.0   \n",
              "2                           0.0                      8192.0   \n",
              "3                        8192.0                         0.0   \n",
              "4                           0.0                         0.0   \n",
              "..                          ...                         ...   \n",
              "115                         0.0                         0.0   \n",
              "116                         0.0                         0.0   \n",
              "117                         0.0                         0.0   \n",
              "118                         0.0                         0.0   \n",
              "119                         0.0                    425984.0   \n",
              "\n",
              "     reads_completed_node1_delta  reads_completed_node2_delta  \\\n",
              "0                            0.0                          0.0   \n",
              "1                            0.0                          1.0   \n",
              "2                            0.0                          0.0   \n",
              "3                            0.0                          2.0   \n",
              "4                            0.0                          0.0   \n",
              "..                           ...                          ...   \n",
              "115                          0.0                          0.0   \n",
              "116                          0.0                          0.0   \n",
              "117                          0.0                          0.0   \n",
              "118                          0.0                          0.0   \n",
              "119                          0.0                          0.0   \n",
              "\n",
              "     reads_completed_node3_delta  worker1_ens192_delta  worker2_ens192_delta  \\\n",
              "0                            3.0             4655872.0             5304403.0   \n",
              "1                            0.0             6787022.0            11630864.0   \n",
              "2                            2.0             4357229.0            20969424.0   \n",
              "3                            0.0             4271090.0            18253010.0   \n",
              "4                            0.0             4388911.0            19194818.0   \n",
              "..                           ...                   ...                   ...   \n",
              "115                          0.0             4208621.0             5785057.0   \n",
              "116                          0.0             4202288.0             4979861.0   \n",
              "117                          0.0             4495483.0             5479759.0   \n",
              "118                          0.0             4471887.0             5366940.0   \n",
              "119                         19.0             4552054.0             6121594.0   \n",
              "\n",
              "     worker3_ens192_delta  produced_logs  \n",
              "0               8135299.0           9769  \n",
              "1              12922413.0          13375  \n",
              "2              28324795.0          11841  \n",
              "3              30140849.0          12263  \n",
              "4              26716078.0          12307  \n",
              "..                    ...            ...  \n",
              "115             9857438.0           1665  \n",
              "116             7215572.0           1678  \n",
              "117             8241919.0           1896  \n",
              "118             7213014.0           1923  \n",
              "119             9237790.0           2017  \n",
              "\n",
              "[120 rows x 46 columns]"
            ],
            "text/html": [
              "\n",
              "\n",
              "  <div id=\"df-ab72f2f3-2d63-4ea8-afcc-c8cb8b4e238b\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>idle_node1_delta</th>\n",
              "      <th>iowait_node1_delta</th>\n",
              "      <th>user_node1_delta</th>\n",
              "      <th>system_node1_delta</th>\n",
              "      <th>irq_node1_delta</th>\n",
              "      <th>softirq_node1_delta</th>\n",
              "      <th>nice_node1_delta</th>\n",
              "      <th>steal_node1_delta</th>\n",
              "      <th>idle_node2_delta</th>\n",
              "      <th>iowait_node2_delta</th>\n",
              "      <th>...</th>\n",
              "      <th>read_bytes_tot_node1_delta</th>\n",
              "      <th>read_bytes_tot_node2_delta</th>\n",
              "      <th>read_bytes_tot_node3_delta</th>\n",
              "      <th>reads_completed_node1_delta</th>\n",
              "      <th>reads_completed_node2_delta</th>\n",
              "      <th>reads_completed_node3_delta</th>\n",
              "      <th>worker1_ens192_delta</th>\n",
              "      <th>worker2_ens192_delta</th>\n",
              "      <th>worker3_ens192_delta</th>\n",
              "      <th>produced_logs</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>440.24</td>\n",
              "      <td>0.45</td>\n",
              "      <td>23.28</td>\n",
              "      <td>8.26</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.02</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>440.24</td>\n",
              "      <td>0.45</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>12288.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4655872.0</td>\n",
              "      <td>5304403.0</td>\n",
              "      <td>8135299.0</td>\n",
              "      <td>9769</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>432.90</td>\n",
              "      <td>0.50</td>\n",
              "      <td>27.34</td>\n",
              "      <td>10.23</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.98</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>432.90</td>\n",
              "      <td>0.50</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4096.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>6787022.0</td>\n",
              "      <td>11630864.0</td>\n",
              "      <td>12922413.0</td>\n",
              "      <td>13375</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>420.56</td>\n",
              "      <td>0.67</td>\n",
              "      <td>35.13</td>\n",
              "      <td>12.70</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.60</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>420.56</td>\n",
              "      <td>0.67</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>8192.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>4357229.0</td>\n",
              "      <td>20969424.0</td>\n",
              "      <td>28324795.0</td>\n",
              "      <td>11841</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>423.56</td>\n",
              "      <td>1.43</td>\n",
              "      <td>31.65</td>\n",
              "      <td>12.28</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.77</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>423.56</td>\n",
              "      <td>1.43</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>8192.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4271090.0</td>\n",
              "      <td>18253010.0</td>\n",
              "      <td>30140849.0</td>\n",
              "      <td>12263</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>424.19</td>\n",
              "      <td>0.52</td>\n",
              "      <td>32.19</td>\n",
              "      <td>11.91</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.64</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>424.19</td>\n",
              "      <td>0.52</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4388911.0</td>\n",
              "      <td>19194818.0</td>\n",
              "      <td>26716078.0</td>\n",
              "      <td>12307</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>115</th>\n",
              "      <td>443.26</td>\n",
              "      <td>0.45</td>\n",
              "      <td>22.57</td>\n",
              "      <td>6.99</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.89</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>443.26</td>\n",
              "      <td>0.45</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4208621.0</td>\n",
              "      <td>5785057.0</td>\n",
              "      <td>9857438.0</td>\n",
              "      <td>1665</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>116</th>\n",
              "      <td>446.77</td>\n",
              "      <td>0.41</td>\n",
              "      <td>19.32</td>\n",
              "      <td>7.41</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.75</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>446.77</td>\n",
              "      <td>0.41</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4202288.0</td>\n",
              "      <td>4979861.0</td>\n",
              "      <td>7215572.0</td>\n",
              "      <td>1678</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>117</th>\n",
              "      <td>419.64</td>\n",
              "      <td>24.85</td>\n",
              "      <td>21.82</td>\n",
              "      <td>7.12</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.85</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>419.64</td>\n",
              "      <td>24.85</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4495483.0</td>\n",
              "      <td>5479759.0</td>\n",
              "      <td>8241919.0</td>\n",
              "      <td>1896</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>118</th>\n",
              "      <td>445.10</td>\n",
              "      <td>0.40</td>\n",
              "      <td>20.26</td>\n",
              "      <td>7.48</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.74</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>445.10</td>\n",
              "      <td>0.40</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4471887.0</td>\n",
              "      <td>5366940.0</td>\n",
              "      <td>7213014.0</td>\n",
              "      <td>1923</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>119</th>\n",
              "      <td>441.12</td>\n",
              "      <td>0.50</td>\n",
              "      <td>24.16</td>\n",
              "      <td>7.78</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.87</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>441.12</td>\n",
              "      <td>0.50</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>425984.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>4552054.0</td>\n",
              "      <td>6121594.0</td>\n",
              "      <td>9237790.0</td>\n",
              "      <td>2017</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>120 rows × 46 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ab72f2f3-2d63-4ea8-afcc-c8cb8b4e238b')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "\n",
              "\n",
              "\n",
              "    <div id=\"df-e1f594cd-66ea-42f2-a342-5578df3030d9\">\n",
              "      <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-e1f594cd-66ea-42f2-a342-5578df3030d9')\"\n",
              "              title=\"Suggest charts.\"\n",
              "              style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "      </button>\n",
              "    </div>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "    background-color: #E8F0FE;\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: #1967D2;\n",
              "    height: 32px;\n",
              "    padding: 0 0 0 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: #E2EBFA;\n",
              "    box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: #174EA6;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "    background-color: #3B4455;\n",
              "    fill: #D2E3FC;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart:hover {\n",
              "    background-color: #434B5C;\n",
              "    box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "    filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "    fill: #FFFFFF;\n",
              "  }\n",
              "</style>\n",
              "\n",
              "    <script>\n",
              "      async function quickchart(key) {\n",
              "        const containerElement = document.querySelector('#' + key);\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      }\n",
              "    </script>\n",
              "\n",
              "      <script>\n",
              "\n",
              "function displayQuickchartButton(domScope) {\n",
              "  let quickchartButtonEl =\n",
              "    domScope.querySelector('#df-e1f594cd-66ea-42f2-a342-5578df3030d9 button.colab-df-quickchart');\n",
              "  quickchartButtonEl.style.display =\n",
              "    google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "}\n",
              "\n",
              "        displayQuickchartButton(document);\n",
              "      </script>\n",
              "      <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ab72f2f3-2d63-4ea8-afcc-c8cb8b4e238b button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ab72f2f3-2d63-4ea8-afcc-c8cb8b4e238b');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test['ts-auth-service'] = test_pods['ts-auth-service-868688bf9c-qp4bp']\n",
        "test['ts-travel-service'] = test_pods['ts-travel-service-59d4847985-sx9xq']\n",
        "test['ts-ui-dashboard'] = test_pods['ts-ui-dashboard-579c4974d4-b5tw2']\n",
        "test['ts-contacts-service'] = test_pods['ts-contacts-service-77fd99dd7f-jp4kg']"
      ],
      "metadata": {
        "id": "FJyklAbSEJ2S"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test['ts-auth-service'].astype(int)\n",
        "test['ts-travel-service'].astype(int)\n",
        "test['ts-ui-dashboard'].astype(int)\n",
        "test['ts-contacts-service'].astype(int)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p9eBInQ2MHyk",
        "outputId": "7399aaf4-228f-4ed6-95aa-9ede5e35cdb1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0      0\n",
              "1      0\n",
              "2      0\n",
              "3      0\n",
              "4      0\n",
              "      ..\n",
              "115    0\n",
              "116    0\n",
              "117    0\n",
              "118    0\n",
              "119    0\n",
              "Name: ts-contacts-service, Length: 120, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_data_scaled = scaler.fit_transform(test)"
      ],
      "metadata": {
        "id": "2q5BqYsAD14t"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "predictions = pd.read_csv('drive/MyDrive/Tesi/Experiment/Data/Test/Refined/predictions.csv', sep=',', index_col=False)\n",
        "#predictions.drop(labels=['Unnamed: 0', 'Unnamed: 0.1'], axis=1, inplace=True)\n",
        "gt = predictions['Gt']"
      ],
      "metadata": {
        "id": "gbIKWMt5jU28"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Autoencoder"
      ],
      "metadata": {
        "id": "Kqbz3u8f_nNc"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from keras import backend as K\n",
        "def root_mean_squared_error(y_true, y_pred):\n",
        "        return K.sqrt(K.mean(K.square(y_pred - y_true)))"
      ],
      "metadata": {
        "id": "buU_0y41S4Sp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the input shape\n",
        "input_shape = (train_data_scaled.shape[1],)  # Number of features (columns)\n",
        "\n",
        "# Define the autoencoder architecture\n",
        "input_layer = Input(shape=input_shape)\n",
        "encoded1 = Dense(32, activation='relu')(input_layer)  # Adjust the number of neurons in the hidden layer as needed\n",
        "encoded2 = Dense(16, activation='relu')(encoded1)\n",
        "encoded3 = Dense(8, activation='relu')(encoded2)\n",
        "decoded3 = Dense(16, activation='relu')(encoded3)\n",
        "decoded2 = Dense(32, activation='relu')(encoded3)\n",
        "decoded1 = Dense(input_shape[0], activation='sigmoid')(decoded2)\n",
        "\n",
        "# Create the autoencoder model\n",
        "autoencoder = Model(input_layer, decoded1)\n",
        "\n",
        "# Compile the model\n",
        "autoencoder.compile(optimizer='adam', loss='mse')\n",
        "\n",
        "# Train the autoencoder\n",
        "autoencoder.fit(train_data_scaled, train_data_scaled, epochs=100, batch_size=64, shuffle=True, validation_data=(val_data1_scaled, val_data1_scaled))\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "FBQSofk3B_IE",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c147077c-9a02-4d85-b83f-93d7545f722b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "71/71 [==============================] - 2s 5ms/step - loss: 1.0395 - val_loss: 0.9049\n",
            "Epoch 2/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.8456 - val_loss: 0.8268\n",
            "Epoch 3/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7896 - val_loss: 0.7899\n",
            "Epoch 4/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7700 - val_loss: 0.7845\n",
            "Epoch 5/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7660 - val_loss: 0.7833\n",
            "Epoch 6/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7642 - val_loss: 0.7852\n",
            "Epoch 7/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7628 - val_loss: 0.7848\n",
            "Epoch 8/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7613 - val_loss: 0.7860\n",
            "Epoch 9/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7594 - val_loss: 0.7848\n",
            "Epoch 10/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7587 - val_loss: 0.7865\n",
            "Epoch 11/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7584 - val_loss: 0.7857\n",
            "Epoch 12/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7582 - val_loss: 0.7861\n",
            "Epoch 13/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7581 - val_loss: 0.7861\n",
            "Epoch 14/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7580 - val_loss: 0.7855\n",
            "Epoch 15/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7579 - val_loss: 0.7865\n",
            "Epoch 16/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7579 - val_loss: 0.7859\n",
            "Epoch 17/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7578 - val_loss: 0.7833\n",
            "Epoch 18/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7578 - val_loss: 0.7846\n",
            "Epoch 19/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7577 - val_loss: 0.7830\n",
            "Epoch 20/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7577 - val_loss: 0.7834\n",
            "Epoch 21/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7577 - val_loss: 0.7831\n",
            "Epoch 22/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7577 - val_loss: 0.7821\n",
            "Epoch 23/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7577 - val_loss: 0.7820\n",
            "Epoch 24/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7577 - val_loss: 0.7813\n",
            "Epoch 25/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7576 - val_loss: 0.7813\n",
            "Epoch 26/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7576 - val_loss: 0.7808\n",
            "Epoch 27/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7576 - val_loss: 0.7811\n",
            "Epoch 28/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7576 - val_loss: 0.7807\n",
            "Epoch 29/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7576 - val_loss: 0.7801\n",
            "Epoch 30/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7576 - val_loss: 0.7795\n",
            "Epoch 31/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7576 - val_loss: 0.7797\n",
            "Epoch 32/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.7576 - val_loss: 0.7798\n",
            "Epoch 33/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7576 - val_loss: 0.7792\n",
            "Epoch 34/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7575 - val_loss: 0.7785\n",
            "Epoch 35/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7575 - val_loss: 0.7790\n",
            "Epoch 36/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7575 - val_loss: 0.7779\n",
            "Epoch 37/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7575 - val_loss: 0.7779\n",
            "Epoch 38/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.7575 - val_loss: 0.7780\n",
            "Epoch 39/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.7575 - val_loss: 0.7770\n",
            "Epoch 40/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7575 - val_loss: 0.7766\n",
            "Epoch 41/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7575 - val_loss: 0.7772\n",
            "Epoch 42/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7575 - val_loss: 0.7768\n",
            "Epoch 43/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7575 - val_loss: 0.7762\n",
            "Epoch 44/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7575 - val_loss: 0.7762\n",
            "Epoch 45/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7574 - val_loss: 0.7755\n",
            "Epoch 46/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7574 - val_loss: 0.7755\n",
            "Epoch 47/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.7574 - val_loss: 0.7751\n",
            "Epoch 48/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7574 - val_loss: 0.7750\n",
            "Epoch 49/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7574 - val_loss: 0.7751\n",
            "Epoch 50/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7574 - val_loss: 0.7745\n",
            "Epoch 51/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7574 - val_loss: 0.7746\n",
            "Epoch 52/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7573 - val_loss: 0.7740\n",
            "Epoch 53/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7573 - val_loss: 0.7739\n",
            "Epoch 54/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7573 - val_loss: 0.7739\n",
            "Epoch 55/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7573 - val_loss: 0.7737\n",
            "Epoch 56/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7572 - val_loss: 0.7733\n",
            "Epoch 57/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7572 - val_loss: 0.7731\n",
            "Epoch 58/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7571 - val_loss: 0.7729\n",
            "Epoch 59/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7571 - val_loss: 0.7730\n",
            "Epoch 60/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7570 - val_loss: 0.7725\n",
            "Epoch 61/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7570 - val_loss: 0.7726\n",
            "Epoch 62/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7570 - val_loss: 0.7727\n",
            "Epoch 63/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7570 - val_loss: 0.7722\n",
            "Epoch 64/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7569 - val_loss: 0.7723\n",
            "Epoch 65/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7569 - val_loss: 0.7724\n",
            "Epoch 66/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7569 - val_loss: 0.7715\n",
            "Epoch 67/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7569 - val_loss: 0.7719\n",
            "Epoch 68/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7569 - val_loss: 0.7717\n",
            "Epoch 69/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7569 - val_loss: 0.7713\n",
            "Epoch 70/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7569 - val_loss: 0.7723\n",
            "Epoch 71/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7569 - val_loss: 0.7717\n",
            "Epoch 72/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7569 - val_loss: 0.7714\n",
            "Epoch 73/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7569 - val_loss: 0.7714\n",
            "Epoch 74/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7569 - val_loss: 0.7713\n",
            "Epoch 75/100\n",
            "71/71 [==============================] - 0s 2ms/step - loss: 0.7568 - val_loss: 0.7714\n",
            "Epoch 76/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.7568 - val_loss: 0.7709\n",
            "Epoch 77/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.7568 - val_loss: 0.7709\n",
            "Epoch 78/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 0.7569 - val_loss: 0.7706\n",
            "Epoch 79/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.7568 - val_loss: 0.7705\n",
            "Epoch 80/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.7568 - val_loss: 0.7709\n",
            "Epoch 81/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.7568 - val_loss: 0.7706\n",
            "Epoch 82/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.7568 - val_loss: 0.7703\n",
            "Epoch 83/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.7568 - val_loss: 0.7702\n",
            "Epoch 84/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 0.7568 - val_loss: 0.7706\n",
            "Epoch 85/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.7568 - val_loss: 0.7700\n",
            "Epoch 86/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.7568 - val_loss: 0.7697\n",
            "Epoch 87/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.7568 - val_loss: 0.7702\n",
            "Epoch 88/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.7568 - val_loss: 0.7698\n",
            "Epoch 89/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 0.7568 - val_loss: 0.7695\n",
            "Epoch 90/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 0.7568 - val_loss: 0.7694\n",
            "Epoch 91/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.7568 - val_loss: 0.7699\n",
            "Epoch 92/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.7568 - val_loss: 0.7699\n",
            "Epoch 93/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 0.7568 - val_loss: 0.7691\n",
            "Epoch 94/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.7568 - val_loss: 0.7748\n",
            "Epoch 95/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.7568 - val_loss: 0.7740\n",
            "Epoch 96/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 0.7568 - val_loss: 0.7740\n",
            "Epoch 97/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.7568 - val_loss: 0.7740\n",
            "Epoch 98/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.7568 - val_loss: 0.7736\n",
            "Epoch 99/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.7568 - val_loss: 0.7737\n",
            "Epoch 100/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 0.7568 - val_loss: 0.7731\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7b8c20ad69e0>"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.plot(autoencoder.history.history['loss'], linewidth=2, label='Train')\n",
        "plt.plot(autoencoder.history.history['val_loss'], linewidth=2, label='Test')\n",
        "plt.legend(loc='upper right')\n",
        "plt.title('Model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "#plt.ylim(ymin=0.70,ymax=1)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "yVa5GPFYH4VQ",
        "outputId": "fb4bf396-ae79-403d-ba5a-f4d4034df024"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "reconstructed_data = autoencoder.predict(val_data2_scaled)\n",
        "mse = np.mean(np.power(val_data2_scaled - reconstructed_data, 2), axis=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9X8SMvOhiVjH",
        "outputId": "b3ff01c8-75e9-4ab6-d291-e7b7ffceff19"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "18/18 [==============================] - 0s 2ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "mse"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UvC0n0CLiprJ",
        "outputId": "b2c0f2e6-83fb-42e4-a413-58b734c0c8af"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 0.72167191,  2.22470929,  1.68766307,  0.25523104,  0.05590143,\n",
              "        0.70431227,  0.59481489,  2.11068197,  0.10895863,  0.81202182,\n",
              "        0.19515968,  0.34772877,  0.52431878,  1.70242528,  1.12470649,\n",
              "        1.2685661 ,  0.09352031,  0.32702485,  0.31691773,  0.10416467,\n",
              "        0.8657886 ,  0.40552248,  0.15402504,  0.46874026,  0.14523219,\n",
              "        0.31348893,  0.85425384,  0.2096397 ,  0.49782489,  0.55661312,\n",
              "        0.17534442,  1.32893657,  0.36014678,  0.44026219,  1.1435633 ,\n",
              "        0.52008543,  4.0275675 ,  0.31066142,  0.46991364,  0.77076003,\n",
              "        1.89306539,  0.30250154,  2.49773549,  0.4315935 ,  0.54444937,\n",
              "        0.49134115,  1.12815832,  0.93440217,  0.80922138,  0.22879299,\n",
              "        0.50624693,  0.16925383,  0.29418924,  0.50115615,  0.55724092,\n",
              "        0.5900457 ,  0.06143927,  0.19644819,  0.80429464,  0.11450404,\n",
              "        0.26098096,  0.1571489 ,  0.17539239,  0.1566097 ,  0.57092078,\n",
              "        0.5083282 ,  0.1070362 ,  0.14880276,  0.05862643,  0.66506729,\n",
              "        1.64078688,  0.94391018,  0.38588996,  0.4832685 ,  0.52573099,\n",
              "        2.7853612 ,  3.24439217,  0.51480115,  0.72059538,  0.41630251,\n",
              "        0.5543673 ,  1.30339798,  0.12627471,  0.36076593,  0.11890486,\n",
              "        0.82411989,  0.06684856,  0.10965421,  0.42018968,  0.36204892,\n",
              "        0.55560995,  0.54419185,  0.50859667,  0.551002  ,  0.33149717,\n",
              "        0.46173375,  0.13586553,  0.62764726,  0.06609544,  0.1307285 ,\n",
              "        0.14442138,  0.20787566,  1.88756762,  0.34603552,  0.35536899,\n",
              "        2.14597133,  0.39122754,  0.38191197,  2.8357114 ,  0.46583792,\n",
              "        0.13754307,  2.16356257,  0.22440883,  0.46366634,  0.53713376,\n",
              "        0.41966469,  0.35644289,  0.16735933,  1.47549986,  0.15225463,\n",
              "        0.13917713,  0.13384125,  0.15721216,  0.71074861,  0.24230341,\n",
              "        0.25966192,  1.10758516,  0.10215676,  0.14451608,  0.58242385,\n",
              "        0.51608985,  1.48797014,  0.12855208,  0.17363927,  0.41927898,\n",
              "        0.34859708,  0.38041575,  0.45432318,  0.41252772,  0.49990311,\n",
              "        0.46938098,  0.35899358,  1.25751217,  0.58342767,  0.59601056,\n",
              "        0.69357375,  3.50894837,  0.6101846 ,  0.1177904 ,  1.31349978,\n",
              "        6.611677  ,  1.0571221 ,  0.30661776,  0.4728798 ,  0.07733018,\n",
              "        0.75830423,  1.41708342,  0.83024129,  0.30535439,  0.94394284,\n",
              "        0.49634119,  0.22547069,  0.28962515,  2.90583768,  0.29484035,\n",
              "        1.40658629,  0.11428322,  0.12533648,  0.58230794,  0.10456079,\n",
              "        0.1797908 ,  1.41443192,  1.12872441,  2.10845509,  0.4561217 ,\n",
              "        0.46419372,  0.11096762,  0.28369059,  0.22496201,  0.28636174,\n",
              "        0.60915041,  0.71860342,  0.22425397,  0.21166542,  0.71184599,\n",
              "        0.68033545,  2.54673674,  0.72877248,  0.33699183,  0.07607773,\n",
              "        0.46008501,  0.20862341,  0.51107818,  0.08474459,  1.25858656,\n",
              "        0.09510784,  0.18002737,  0.60743421,  1.05463852,  0.30737196,\n",
              "        0.38137545,  0.12356462,  0.30695734,  0.63366113,  0.39984486,\n",
              "        0.83428402,  0.3612662 ,  0.17658138,  0.45033444,  0.09666777,\n",
              "        0.21681836,  0.82956272,  0.92313935,  0.17590318,  0.27662188,\n",
              "        0.74435718,  0.32631744,  0.20389887,  1.02451305,  0.84947713,\n",
              "        0.24314306,  0.14638638,  0.31832148,  0.66010784,  0.44628556,\n",
              "        1.04655803,  0.53673824,  0.10023011,  0.17496513,  0.39757955,\n",
              "        0.44206661,  0.05320218,  0.80362278,  0.57707653,  0.14433285,\n",
              "        0.43798367,  0.10912692,  0.3332374 ,  0.19019094,  0.80957917,\n",
              "        0.30651763,  0.59407869,  0.70897749,  0.30860907,  0.35541475,\n",
              "        0.46894584,  0.0972344 ,  2.79666682, 26.22120323,  0.71264811,\n",
              "        0.17118673,  0.31855272,  1.08133298,  0.13216574,  0.32030077,\n",
              "        0.20617929,  0.23686588,  0.3242101 ,  0.83127259,  0.09265725,\n",
              "        1.28015896,  1.30649619,  0.22449547,  0.55542347,  0.22470796,\n",
              "        0.05501831,  0.32854562,  0.35785677,  0.60727811,  0.35322638,\n",
              "        1.38029957,  0.11582222,  2.67970985,  0.11847447,  0.17131486,\n",
              "        0.2304779 ,  0.66940669,  0.50710881,  0.33529405,  0.72568886,\n",
              "        0.46834475,  2.68913735,  0.5268708 ,  0.7000217 ,  0.93926492,\n",
              "        0.16628961,  0.16497351,  0.0564578 ,  0.45681731,  0.37937742,\n",
              "        5.17987453,  0.97443255,  2.28474877,  0.11384169,  3.83878364,\n",
              "        0.12277478,  0.08617723,  0.41221936,  0.41887907,  0.41401261,\n",
              "        0.43928474,  0.20958051,  0.20492821,  0.60045451,  0.45094746,\n",
              "        0.44358695,  0.74616807,  1.7648594 ,  0.36144861,  2.83079623,\n",
              "        0.43979836,  0.3249695 ,  0.59971716,  0.36134019,  0.47117297,\n",
              "        0.73118085,  0.1673264 ,  0.2132875 ,  0.72722275,  0.15609423,\n",
              "        1.07919526,  0.12143176,  0.09191441,  0.08481184,  0.3435482 ,\n",
              "        0.20138848,  0.392237  ,  0.27591099,  0.33981607,  0.09612288,\n",
              "        0.22131897,  0.60344233,  0.39230676,  0.66054778,  0.50085715,\n",
              "        0.11054249,  0.34260061,  2.23907516,  0.27005705,  0.16721837,\n",
              "        0.13279789,  3.11100699,  0.41309117,  0.28018747,  0.1445794 ,\n",
              "        1.93084738,  0.67751105,  0.20729042,  0.23470135,  0.45207786,\n",
              "        0.43776545,  0.55535004,  0.07920111,  0.58127697,  0.44455634,\n",
              "        0.21574204,  0.23890011,  0.16277801,  0.18672445,  5.12090388,\n",
              "        0.69382922,  0.09864219,  1.96815458,  2.17882095,  0.14043209,\n",
              "        0.2188775 ,  1.00051387, 32.76920509,  0.14161367,  0.51642433,\n",
              "        0.07423111,  0.54782389,  0.55910214,  0.12083446,  0.66340062,\n",
              "        0.81227397,  2.05330738,  0.69354817,  0.39031929,  0.13712368,\n",
              "        0.85291937,  0.13030726,  0.31377127,  0.64876649,  1.10584118,\n",
              "        0.39199459,  0.10493966,  0.24970901,  0.12957015,  0.19850005,\n",
              "        1.12430508,  0.26906424,  2.31868602,  0.48905555,  0.53541606,\n",
              "        0.26716377,  0.19084021,  0.80342447,  0.66619077,  0.77251839,\n",
              "        0.23011   ,  0.51918233,  1.17251329,  0.47736941,  0.47880653,\n",
              "        0.48435989,  0.15482869,  0.60067711,  0.28620204,  0.62544448,\n",
              "        0.14828702,  0.92408816,  0.28852874,  0.41035024,  1.05951234,\n",
              "        0.57121109,  1.06992884,  0.09005767,  1.36401523,  0.6341947 ,\n",
              "        0.13502978,  0.46575845,  0.22601047,  0.12467154,  0.26105217,\n",
              "        0.11542612,  0.56481856,  0.31315619,  0.23164567,  0.29287571,\n",
              "        0.80642164,  0.66911768,  1.04271979,  0.12793166,  0.29421995,\n",
              "        0.474155  ,  0.34936043,  0.12752032,  0.17656094,  0.24039892,\n",
              "        0.23179384,  1.69449957,  0.06204231,  0.63820697,  0.12015859,\n",
              "        0.79143918,  0.70921134,  1.14424076,  0.19486608,  0.13662105,\n",
              "        0.07340886,  0.21422794,  2.72363282,  0.11961114,  4.41512749,\n",
              "        0.41118471,  0.16475511,  0.15800784,  0.98405126,  0.17626227,\n",
              "        0.18749935,  0.40684926,  0.16527286,  0.13164246,  0.58797948,\n",
              "        0.51477244,  0.17892994,  0.20284106,  0.13891866,  0.15013532,\n",
              "        0.22761368,  0.17141381,  0.20610291,  0.52998091,  0.16274672,\n",
              "        0.09439597,  0.10373065,  0.45074964,  0.66564153,  0.54131919,\n",
              "        0.27559271,  0.5153126 ,  0.61962943,  1.75336631,  0.21239457,\n",
              "        0.15851061,  2.81424409, 13.55420835,  0.57515457,  0.4155397 ,\n",
              "        0.10562707,  0.58982489,  0.12536857,  0.07219965,  0.07636654,\n",
              "        0.7283993 ,  0.71229411,  1.02921679,  0.62298442,  0.208774  ,\n",
              "        0.69676432,  0.6947947 ,  0.7441571 ,  0.35798707,  0.22848726,\n",
              "        0.2188513 ,  0.50506977,  0.16215685,  0.67577458,  0.04492035,\n",
              "        0.15792211,  0.53239105,  0.43029644,  0.10702931,  2.87687773,\n",
              "        0.18654655,  0.51367437,  0.91371376,  0.54226988,  0.72656906,\n",
              "        0.32956304,  0.48055112,  1.53219076,  0.51211202,  0.38968194,\n",
              "        0.12933549,  0.13386302,  0.6820697 ,  0.62604927,  0.60410821,\n",
              "        0.09438759,  0.29421809,  0.18684086,  0.72359158,  0.39781992,\n",
              "        0.42329414,  0.45906881,  0.20662388,  0.16847471,  0.5424672 ,\n",
              "        0.50242123,  0.46539933,  1.40108761,  1.29756427,  0.20257646,\n",
              "        0.73551017,  0.70452432,  0.1709679 ,  0.45858971,  0.21192756,\n",
              "        0.49703736,  2.00230621,  0.06493642,  0.51403889,  0.4349075 ,\n",
              "        0.04207559,  0.7932648 ,  0.3877949 ,  0.4450397 ,  0.98256431,\n",
              "        0.40720277,  0.1215292 ,  0.27340047,  0.16799661,  0.1971645 ,\n",
              "        0.51953936,  0.10503571])"
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.hist(mse)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "s7bV4sjDUEnc",
        "outputId": "43ddad03-c391-48c4-904a-6c9c3f7f6bc9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "stderr = np.std(mse)/np.sqrt(len(mse))\n",
        "stderr"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "krm0DjURsgU4",
        "outputId": "84685d75-9ed4-4450-8239-59507eed7be7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.08172505140578516"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "np.percentile(mse, 90)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Z6s288uGUMMY",
        "outputId": "a7969c67-f403-4e2a-cddc-bef1b24828ae"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1.3046372669760642"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.mean(mse) + np.std(mse)\n",
        "threshold"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Lq5X1MNwJPly",
        "outputId": "7335b947-0d43-4e3e-8d6b-4cac8fe78f59"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2.6775913024982727"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Use the trained autoencoder for anomaly detection\n",
        "#test_data_scaled = scaler.fit_transform(test)\n",
        "reconstructed_data = autoencoder.predict(test_data_scaled)\n",
        "mse = np.mean(np.power(test_data_scaled - reconstructed_data, 2), axis=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SKduxMxE_xng",
        "outputId": "b970c237-339c-4804-9ba9-d2aaf8510e35"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4/4 [==============================] - 0s 3ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(mse)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "do4JC0GskfkK",
        "outputId": "fb1508a3-8b39-4864-d591-c6f378a125b9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[0.143581   2.35816278 3.4802394  3.09586573 2.47280806 1.52474172\n",
            " 1.45110623 1.45304326 1.71584435 1.55000153 1.59899279 2.32540533\n",
            " 1.66216177 2.38195379 1.82780452 1.8050744  1.51635084 0.81948226\n",
            " 0.25931983 0.12900202 0.38794719 0.5963485  0.40848445 0.39119382\n",
            " 0.28823783 0.15714815 0.1904815  0.13398169 0.11564828 0.20888397\n",
            " 0.68871159 0.16334019 0.13457004 1.05865183 0.08284605 0.11466114\n",
            " 0.33376784 0.37347738 0.11590969 0.0728435  0.12665362 0.12953997\n",
            " 0.08346156 0.21174279 0.18403383 0.72054342 0.15256635 0.1624694\n",
            " 0.19894911 0.14102544 1.68493094 0.47480772 0.27606151 0.67017306\n",
            " 0.22881271 0.16123326 4.48139548 0.38667774 1.80566327 0.15399637\n",
            " 0.23052441 0.2065895  0.11242218 0.19602992 0.53667467 0.17702882\n",
            " 0.28959801 0.21147864 0.22255375 0.35460898 0.66101282 0.09524715\n",
            " 0.20012475 2.09669594 0.22346924 0.3047977  0.16445965 0.74634184\n",
            " 0.15282568 0.13815512 0.25278537 0.22517299 0.28133501 0.2584408\n",
            " 0.43674978 0.31571155 0.3439709  0.33906412 0.33126617 0.46473368\n",
            " 0.54266057 1.02346809 0.48917046 0.43096904 0.4521344  0.20951995\n",
            " 0.27629606 0.80178229 0.71387448 0.23802682 0.13389966 0.18927408\n",
            " 5.74799427 0.27740547 0.55950223 0.38762427 1.60267955 0.08781871\n",
            " 0.15120289 0.1338094  0.14222034 0.29220121 0.2587483  0.25401897\n",
            " 0.3358145  0.38374909 0.56095441 5.1036843  0.798017   4.68942111]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred = [1 if score > 1 else 0 for score in mse]"
      ],
      "metadata": {
        "id": "Ocz8-DjuNkoH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "pred"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zonS6nS8fAT6",
        "outputId": "0cfb159f-81b5-47f1-eae4-044923b8e9f3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 1]"
            ]
          },
          "metadata": {},
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(gt, pred, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w3Da-1VMT8rn",
        "outputId": "58a793cb-6b8d-4b55-9437-add968bb375b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.68      0.90      0.78        71\n",
            "      attack       0.73      0.39      0.51        49\n",
            "\n",
            "    accuracy                           0.69       120\n",
            "   macro avg       0.71      0.64      0.64       120\n",
            "weighted avg       0.70      0.69      0.67       120\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Overcomplete Autoencoder"
      ],
      "metadata": {
        "id": "EnuhImn6DgLi"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras.models import Model\n",
        "from tensorflow.keras.layers import Input, Dense, Dropout, BatchNormalization, concatenate\n",
        "\n",
        "def UNet(input_dim, hidden_dim, output_dim):\n",
        "    inputs = Input(shape=(input_dim,))\n",
        "\n",
        "    # Encoder\n",
        "    encoder1 = Dense(hidden_dim, activation='relu')(inputs)\n",
        "    encoder1 = Dropout(0.1)(encoder1)\n",
        "    encoder1 = BatchNormalization()(encoder1)\n",
        "\n",
        "    encoder2 = Dense(hidden_dim, activation='relu')(encoder1)\n",
        "    encoder2 = Dropout(0.1)(encoder2)\n",
        "    encoder2 = BatchNormalization()(encoder2)\n",
        "\n",
        "    # Decoder\n",
        "    decoder2 = Dense(hidden_dim, activation='relu')(encoder2)\n",
        "    decoder2 = Dropout(0.1)(decoder2)\n",
        "    decoder2 = BatchNormalization()(decoder2)\n",
        "\n",
        "    decoder1 = Dense(hidden_dim, activation='relu')(decoder2)\n",
        "    decoder1 = Dropout(0.1)(decoder1)\n",
        "    decoder1 = BatchNormalization()(decoder1)\n",
        "\n",
        "    # Output\n",
        "    output = Dense(output_dim, activation='sigmoid')(decoder1)\n",
        "\n",
        "    model = Model(inputs=inputs, outputs=output)\n",
        "    return model\n",
        "\n",
        "# Define your dataset and data preprocessing\n",
        "# ...\n",
        "\n",
        "# Define hyperparameters\n",
        "input_dim = train_data.shape[1]\n",
        "hidden_dim = 128\n",
        "output_dim = train_data.shape[1]\n",
        "learning_rate = 0.001\n",
        "batch_size = 64\n",
        "num_epochs = 100\n",
        "\n",
        "# Instantiate the UNet model\n",
        "model = UNet(input_dim, hidden_dim, output_dim)\n",
        "\n",
        "# Define loss function and optimizer\n",
        "loss_fn = tf.keras.losses.MeanSquaredError()\n",
        "optimizer = tf.keras.optimizers.Adam(learning_rate=learning_rate)\n",
        "\n",
        "# Compile the model\n",
        "model.compile(optimizer=optimizer, loss=loss_fn)\n",
        "\n",
        "# Train the model\n",
        "model.fit(train_data_scaled, train_data_scaled, epochs=num_epochs, batch_size=batch_size, shuffle=True, validation_data=(val_data1_scaled, val_data1_scaled))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1BAPmioYDhxo",
        "outputId": "4db90dbe-6259-4379-d41e-535c73a3d65e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "71/71 [==============================] - 4s 9ms/step - loss: 0.1810 - val_loss: 0.1729\n",
            "Epoch 2/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0521 - val_loss: 0.2940\n",
            "Epoch 3/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 0.0087 - val_loss: 0.3298\n",
            "Epoch 4/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 0.0044 - val_loss: 0.3402\n",
            "Epoch 5/100\n",
            "71/71 [==============================] - 1s 14ms/step - loss: 0.0032 - val_loss: 0.3457\n",
            "Epoch 6/100\n",
            "71/71 [==============================] - 1s 14ms/step - loss: 0.0027 - val_loss: 0.3456\n",
            "Epoch 7/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 0.0024 - val_loss: 0.3463\n",
            "Epoch 8/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0020 - val_loss: 0.3069\n",
            "Epoch 9/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0019 - val_loss: 0.2493\n",
            "Epoch 10/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0018 - val_loss: 0.2140\n",
            "Epoch 11/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0018 - val_loss: 0.2433\n",
            "Epoch 12/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0015 - val_loss: 0.2266\n",
            "Epoch 13/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0015 - val_loss: 0.2107\n",
            "Epoch 14/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0014 - val_loss: 0.1996\n",
            "Epoch 15/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0013 - val_loss: 0.2309\n",
            "Epoch 16/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 0.0014 - val_loss: 0.2689\n",
            "Epoch 17/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 0.0013 - val_loss: 0.2086\n",
            "Epoch 18/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0012 - val_loss: 0.2794\n",
            "Epoch 19/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0012 - val_loss: 0.2529\n",
            "Epoch 20/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 0.0011 - val_loss: 0.2025\n",
            "Epoch 21/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 0.0011 - val_loss: 0.2134\n",
            "Epoch 22/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 0.0011 - val_loss: 0.2554\n",
            "Epoch 23/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0011 - val_loss: 0.2095\n",
            "Epoch 24/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 9.9903e-04 - val_loss: 0.1616\n",
            "Epoch 25/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 9.8078e-04 - val_loss: 0.1888\n",
            "Epoch 26/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 9.8966e-04 - val_loss: 0.1534\n",
            "Epoch 27/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 9.7194e-04 - val_loss: 0.2472\n",
            "Epoch 28/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 9.7233e-04 - val_loss: 0.2335\n",
            "Epoch 29/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 9.6937e-04 - val_loss: 0.1777\n",
            "Epoch 30/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0010 - val_loss: 0.1771\n",
            "Epoch 31/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 9.4298e-04 - val_loss: 0.1298\n",
            "Epoch 32/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 8.5149e-04 - val_loss: 0.1201\n",
            "Epoch 33/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 9.0751e-04 - val_loss: 0.2562\n",
            "Epoch 34/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 9.0681e-04 - val_loss: 0.2084\n",
            "Epoch 35/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 8.7097e-04 - val_loss: 0.1174\n",
            "Epoch 36/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 8.3691e-04 - val_loss: 0.1180\n",
            "Epoch 37/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 8.2253e-04 - val_loss: 0.1264\n",
            "Epoch 38/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 7.8489e-04 - val_loss: 0.1133\n",
            "Epoch 39/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 8.3884e-04 - val_loss: 0.2775\n",
            "Epoch 40/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 8.3757e-04 - val_loss: 0.2321\n",
            "Epoch 41/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 8.3943e-04 - val_loss: 0.1990\n",
            "Epoch 42/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 7.5447e-04 - val_loss: 0.1244\n",
            "Epoch 43/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 7.4939e-04 - val_loss: 0.1214\n",
            "Epoch 44/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 7.3712e-04 - val_loss: 0.1339\n",
            "Epoch 45/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 7.2000e-04 - val_loss: 0.1349\n",
            "Epoch 46/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 7.2512e-04 - val_loss: 0.1382\n",
            "Epoch 47/100\n",
            "71/71 [==============================] - 1s 14ms/step - loss: 7.8779e-04 - val_loss: 0.1694\n",
            "Epoch 48/100\n",
            "71/71 [==============================] - 1s 15ms/step - loss: 7.0415e-04 - val_loss: 0.1755\n",
            "Epoch 49/100\n",
            "71/71 [==============================] - 1s 14ms/step - loss: 7.3032e-04 - val_loss: 0.1025\n",
            "Epoch 50/100\n",
            "71/71 [==============================] - 1s 14ms/step - loss: 6.9597e-04 - val_loss: 0.1034\n",
            "Epoch 51/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 7.0456e-04 - val_loss: 0.1024\n",
            "Epoch 52/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 7.4327e-04 - val_loss: 0.1093\n",
            "Epoch 53/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 6.6503e-04 - val_loss: 0.1068\n",
            "Epoch 54/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 6.7736e-04 - val_loss: 0.1076\n",
            "Epoch 55/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 7.0696e-04 - val_loss: 0.1084\n",
            "Epoch 56/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 6.9733e-04 - val_loss: 0.1080\n",
            "Epoch 57/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 6.5822e-04 - val_loss: 0.1081\n",
            "Epoch 58/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 6.5596e-04 - val_loss: 0.1052\n",
            "Epoch 59/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 6.2864e-04 - val_loss: 0.1052\n",
            "Epoch 60/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 6.3705e-04 - val_loss: 0.1080\n",
            "Epoch 61/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 6.1445e-04 - val_loss: 0.1087\n",
            "Epoch 62/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 6.3112e-04 - val_loss: 0.1093\n",
            "Epoch 63/100\n",
            "71/71 [==============================] - 1s 14ms/step - loss: 5.8398e-04 - val_loss: 0.1077\n",
            "Epoch 64/100\n",
            "71/71 [==============================] - 1s 16ms/step - loss: 6.1463e-04 - val_loss: 0.1072\n",
            "Epoch 65/100\n",
            "71/71 [==============================] - 1s 17ms/step - loss: 5.9711e-04 - val_loss: 0.1087\n",
            "Epoch 66/100\n",
            "71/71 [==============================] - 2s 26ms/step - loss: 6.1278e-04 - val_loss: 0.1085\n",
            "Epoch 67/100\n",
            "71/71 [==============================] - 2s 21ms/step - loss: 6.0022e-04 - val_loss: 0.1069\n",
            "Epoch 68/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 5.8218e-04 - val_loss: 0.1083\n",
            "Epoch 69/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 6.0290e-04 - val_loss: 0.1092\n",
            "Epoch 70/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 6.3329e-04 - val_loss: 0.2861\n",
            "Epoch 71/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 5.9873e-04 - val_loss: 0.1044\n",
            "Epoch 72/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 6.3393e-04 - val_loss: 0.1753\n",
            "Epoch 73/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 6.0988e-04 - val_loss: 0.1679\n",
            "Epoch 74/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 6.2366e-04 - val_loss: 0.1096\n",
            "Epoch 75/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.7923e-04 - val_loss: 0.1062\n",
            "Epoch 76/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.7488e-04 - val_loss: 0.1079\n",
            "Epoch 77/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.4124e-04 - val_loss: 0.1059\n",
            "Epoch 78/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.5375e-04 - val_loss: 0.1043\n",
            "Epoch 79/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.7530e-04 - val_loss: 0.1060\n",
            "Epoch 80/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.2689e-04 - val_loss: 0.1053\n",
            "Epoch 81/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.3362e-04 - val_loss: 0.1071\n",
            "Epoch 82/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 5.0052e-04 - val_loss: 0.1070\n",
            "Epoch 83/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 5.0260e-04 - val_loss: 0.1067\n",
            "Epoch 84/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 5.4338e-04 - val_loss: 0.1078\n",
            "Epoch 85/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 5.0979e-04 - val_loss: 0.1077\n",
            "Epoch 86/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 5.0994e-04 - val_loss: 0.1064\n",
            "Epoch 87/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 5.0249e-04 - val_loss: 0.1096\n",
            "Epoch 88/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 5.0621e-04 - val_loss: 0.1134\n",
            "Epoch 89/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 4.7755e-04 - val_loss: 0.1139\n",
            "Epoch 90/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.9103e-04 - val_loss: 0.1100\n",
            "Epoch 91/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 5.0408e-04 - val_loss: 0.1103\n",
            "Epoch 92/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.8464e-04 - val_loss: 0.1116\n",
            "Epoch 93/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 4.5940e-04 - val_loss: 0.1151\n",
            "Epoch 94/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.6393e-04 - val_loss: 0.1151\n",
            "Epoch 95/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.7612e-04 - val_loss: 0.1122\n",
            "Epoch 96/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.7833e-04 - val_loss: 0.1132\n",
            "Epoch 97/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.5348e-04 - val_loss: 0.1108\n",
            "Epoch 98/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.8944e-04 - val_loss: 0.1133\n",
            "Epoch 99/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.6348e-04 - val_loss: 0.1107\n",
            "Epoch 100/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.8611e-04 - val_loss: 0.1118\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7ff23e642c50>"
            ]
          },
          "metadata": {},
          "execution_count": 270
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras.models import Model\n",
        "from tensorflow.keras.layers import Input, Dense, Dropout, BatchNormalization, concatenate\n",
        "\n",
        "def UNet(input_dim, hidden_dim, output_dim):\n",
        "    inputs = Input(shape=(input_dim,))\n",
        "\n",
        "    # Encoder\n",
        "    encoder1 = Dense(hidden_dim, activation='relu')(inputs)\n",
        "    encoder1 = Dropout(0.1)(encoder1)\n",
        "    encoder1 = BatchNormalization()(encoder1)\n",
        "\n",
        "    # Output\n",
        "    output = Dense(output_dim, activation='sigmoid')(encoder1)\n",
        "\n",
        "    model = Model(inputs=inputs, outputs=output)\n",
        "    return model\n",
        "\n",
        "# Define your dataset and data preprocessing\n",
        "# ...\n",
        "\n",
        "# Define hyperparameters\n",
        "input_dim = train_data.shape[1]\n",
        "hidden_dim = 64\n",
        "output_dim = train_data.shape[1]\n",
        "learning_rate = 0.001\n",
        "batch_size = 64\n",
        "num_epochs = 100\n",
        "\n",
        "# Instantiate the UNet model\n",
        "model = UNet(input_dim, hidden_dim, output_dim)\n",
        "\n",
        "# Define loss function and optimizer\n",
        "loss_fn = tf.keras.losses.MeanSquaredError()\n",
        "optimizer = tf.keras.optimizers.Adam(learning_rate=learning_rate)\n",
        "\n",
        "# Compile the model\n",
        "model.compile(optimizer=optimizer, loss=loss_fn)\n",
        "\n",
        "# Train the model\n",
        "model.fit(train_data_scaled, train_data_scaled, epochs=num_epochs, batch_size=batch_size, shuffle=True, validation_data=(val_data1_scaled, val_data1_scaled))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FDnK9wv9zwe5",
        "outputId": "9218edf8-d5d5-4d06-99e8-669a7adc4175"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "71/71 [==============================] - 1s 6ms/step - loss: 0.1646 - val_loss: 0.1756\n",
            "Epoch 2/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0527 - val_loss: 0.2894\n",
            "Epoch 3/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0110 - val_loss: 0.3338\n",
            "Epoch 4/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0050 - val_loss: 0.3449\n",
            "Epoch 5/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0032 - val_loss: 0.3452\n",
            "Epoch 6/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0024 - val_loss: 0.3471\n",
            "Epoch 7/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0020 - val_loss: 0.3304\n",
            "Epoch 8/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0017 - val_loss: 0.3211\n",
            "Epoch 9/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0015 - val_loss: 0.3094\n",
            "Epoch 10/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0015 - val_loss: 0.2903\n",
            "Epoch 11/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0014 - val_loss: 0.2808\n",
            "Epoch 12/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0013 - val_loss: 0.2768\n",
            "Epoch 13/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0012 - val_loss: 0.2676\n",
            "Epoch 14/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0012 - val_loss: 0.2666\n",
            "Epoch 15/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0011 - val_loss: 0.2677\n",
            "Epoch 16/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 0.0011 - val_loss: 0.2650\n",
            "Epoch 17/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.0011 - val_loss: 0.2446\n",
            "Epoch 18/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.0010 - val_loss: 0.2425\n",
            "Epoch 19/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.0010 - val_loss: 0.2451\n",
            "Epoch 20/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.0010 - val_loss: 0.2500\n",
            "Epoch 21/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 9.9176e-04 - val_loss: 0.2561\n",
            "Epoch 22/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 0.0010 - val_loss: 0.2384\n",
            "Epoch 23/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 9.8458e-04 - val_loss: 0.2470\n",
            "Epoch 24/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 9.6624e-04 - val_loss: 0.2216\n",
            "Epoch 25/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 9.4918e-04 - val_loss: 0.2258\n",
            "Epoch 26/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 9.4616e-04 - val_loss: 0.2344\n",
            "Epoch 27/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 9.3874e-04 - val_loss: 0.2323\n",
            "Epoch 28/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 9.4619e-04 - val_loss: 0.2309\n",
            "Epoch 29/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.9509e-04 - val_loss: 0.2348\n",
            "Epoch 30/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.9719e-04 - val_loss: 0.2361\n",
            "Epoch 31/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.8138e-04 - val_loss: 0.2278\n",
            "Epoch 32/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 9.1385e-04 - val_loss: 0.2283\n",
            "Epoch 33/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.8762e-04 - val_loss: 0.2304\n",
            "Epoch 34/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.7171e-04 - val_loss: 0.2237\n",
            "Epoch 35/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 8.7235e-04 - val_loss: 0.2260\n",
            "Epoch 36/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.5757e-04 - val_loss: 0.2261\n",
            "Epoch 37/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.5164e-04 - val_loss: 0.2269\n",
            "Epoch 38/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.3928e-04 - val_loss: 0.2213\n",
            "Epoch 39/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.2475e-04 - val_loss: 0.2153\n",
            "Epoch 40/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.4779e-04 - val_loss: 0.2136\n",
            "Epoch 41/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.5093e-04 - val_loss: 0.2120\n",
            "Epoch 42/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.3804e-04 - val_loss: 0.1852\n",
            "Epoch 43/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.8454e-04 - val_loss: 0.1664\n",
            "Epoch 44/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.1803e-04 - val_loss: 0.1669\n",
            "Epoch 45/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 9.0259e-04 - val_loss: 0.2372\n",
            "Epoch 46/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.2357e-04 - val_loss: 0.2341\n",
            "Epoch 47/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.1553e-04 - val_loss: 0.2181\n",
            "Epoch 48/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.3171e-04 - val_loss: 0.2207\n",
            "Epoch 49/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 7.7412e-04 - val_loss: 0.2213\n",
            "Epoch 50/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 8.3483e-04 - val_loss: 0.2048\n",
            "Epoch 51/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 7.7946e-04 - val_loss: 0.2068\n",
            "Epoch 52/100\n",
            "71/71 [==============================] - 0s 3ms/step - loss: 7.9675e-04 - val_loss: 0.2139\n",
            "Epoch 53/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 8.1885e-04 - val_loss: 0.2135\n",
            "Epoch 54/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 8.3835e-04 - val_loss: 0.2145\n",
            "Epoch 55/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 7.8228e-04 - val_loss: 0.2168\n",
            "Epoch 56/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.9586e-04 - val_loss: 0.2166\n",
            "Epoch 57/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.7920e-04 - val_loss: 0.2178\n",
            "Epoch 58/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 7.7606e-04 - val_loss: 0.2150\n",
            "Epoch 59/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 8.1064e-04 - val_loss: 0.2163\n",
            "Epoch 60/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.9908e-04 - val_loss: 0.1867\n",
            "Epoch 61/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 7.9204e-04 - val_loss: 0.2140\n",
            "Epoch 62/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 8.4315e-04 - val_loss: 0.2111\n",
            "Epoch 63/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 8.1054e-04 - val_loss: 0.2111\n",
            "Epoch 64/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.8203e-04 - val_loss: 0.2113\n",
            "Epoch 65/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 8.1372e-04 - val_loss: 0.2124\n",
            "Epoch 66/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.8133e-04 - val_loss: 0.2112\n",
            "Epoch 67/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 7.7158e-04 - val_loss: 0.2108\n",
            "Epoch 68/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 8.1092e-04 - val_loss: 0.2103\n",
            "Epoch 69/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 8.3618e-04 - val_loss: 0.2102\n",
            "Epoch 70/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 8.1588e-04 - val_loss: 0.2128\n",
            "Epoch 71/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 8.2108e-04 - val_loss: 0.2127\n",
            "Epoch 72/100\n",
            "71/71 [==============================] - 1s 14ms/step - loss: 7.9097e-04 - val_loss: 0.2120\n",
            "Epoch 73/100\n",
            "71/71 [==============================] - 1s 16ms/step - loss: 7.9229e-04 - val_loss: 0.2133\n",
            "Epoch 74/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 8.1133e-04 - val_loss: 0.2144\n",
            "Epoch 75/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 7.9998e-04 - val_loss: 0.2132\n",
            "Epoch 76/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 8.0691e-04 - val_loss: 0.2157\n",
            "Epoch 77/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 7.4581e-04 - val_loss: 0.2145\n",
            "Epoch 78/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 8.0615e-04 - val_loss: 0.2171\n",
            "Epoch 79/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 7.7844e-04 - val_loss: 0.1990\n",
            "Epoch 80/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 7.9755e-04 - val_loss: 0.1986\n",
            "Epoch 81/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 7.5104e-04 - val_loss: 0.2008\n",
            "Epoch 82/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 7.7443e-04 - val_loss: 0.1757\n",
            "Epoch 83/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 7.8017e-04 - val_loss: 0.1751\n",
            "Epoch 84/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 7.5542e-04 - val_loss: 0.1766\n",
            "Epoch 85/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 7.7017e-04 - val_loss: 0.1780\n",
            "Epoch 86/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 7.8471e-04 - val_loss: 0.1773\n",
            "Epoch 87/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 7.4262e-04 - val_loss: 0.1773\n",
            "Epoch 88/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 7.4644e-04 - val_loss: 0.1792\n",
            "Epoch 89/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.7125e-04 - val_loss: 0.1784\n",
            "Epoch 90/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 7.9183e-04 - val_loss: 0.1747\n",
            "Epoch 91/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 7.8949e-04 - val_loss: 0.1614\n",
            "Epoch 92/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 7.2831e-04 - val_loss: 0.1629\n",
            "Epoch 93/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 7.8306e-04 - val_loss: 0.1629\n",
            "Epoch 94/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 7.5089e-04 - val_loss: 0.1647\n",
            "Epoch 95/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 7.4001e-04 - val_loss: 0.1649\n",
            "Epoch 96/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.8701e-04 - val_loss: 0.1662\n",
            "Epoch 97/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.5636e-04 - val_loss: 0.1680\n",
            "Epoch 98/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.6754e-04 - val_loss: 0.1675\n",
            "Epoch 99/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.4867e-04 - val_loss: 0.1671\n",
            "Epoch 100/100\n",
            "71/71 [==============================] - 0s 4ms/step - loss: 7.7394e-04 - val_loss: 0.1667\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f31bd257d30>"
            ]
          },
          "metadata": {},
          "execution_count": 46
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.plot(model.history.history['loss'], linewidth=2, label='Train')\n",
        "plt.plot(model.history.history['val_loss'], linewidth=2, label='Val')\n",
        "plt.legend(loc='upper right')\n",
        "plt.title('Model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "#plt.ylim(ymin=0.70,ymax=1)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "l33v6bwVN9Up",
        "outputId": "df05cb6b-f54e-46a7-db0e-5265620f4d61"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkAAAAHHCAYAAABXx+fLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABoeklEQVR4nO3dd3gU1foH8O/uJtn0AiENAqFJ6AECMYiCEgmICAoKiILRn9wL4gVj5aqA7QZQEQvCVS9FBUVUUFFAjASlN5HeFAhtU4B00nbn98fJlglJSNnd2ex+P8+zT2Znzs6emZR9c857zlFJkiSBiIiIyIWola4AERERkb0xACIiIiKXwwCIiIiIXA4DICIiInI5DICIiIjI5TAAIiIiIpfDAIiIiIhcDgMgIiIicjkMgIiIiMjlMAAiokZPpVJh1qxZdX7dmTNnoFKpsHTp0hrLpaWlQaVSIS0trV71IyLHwwCIiKxi6dKlUKlUUKlU2LJly3XHJUlCZGQkVCoV7r77bgVqSERkxgCIiKzK09MTK1asuG7/5s2bcf78eWi1WgVqRUQkxwCIiKzqrrvuwqpVq1BeXi7bv2LFCvTq1QthYWEK1YyIyIwBEBFZ1dixY3H58mVs3LjRtK+0tBRff/01HnzwwSpfU1hYiKeffhqRkZHQarXo0KED3nrrLUiSJCtXUlKCp556Cs2aNYOfnx/uuecenD9/vspzXrhwAY8++ihCQ0Oh1WrRuXNnLF682HoXCmDVqlXo1asXvLy8EBwcjIceeggXLlyQldHpdEhKSkKLFi2g1WoRHh6O4cOH48yZM6Yye/bsQWJiIoKDg+Hl5YXWrVvj0UcftWpdiUjOTekKEJFziYqKQnx8PL744gsMGTIEALBu3Trk5uZizJgxeO+992TlJUnCPffcg02bNuGxxx5DTEwMNmzYgGeffRYXLlzAO++8Yyr7f//3f/j888/x4IMPom/fvvj1118xdOjQ6+qQkZGBm2++GSqVClOmTEGzZs2wbt06PPbYY8jLy8O0adMafJ1Lly5FUlISevfujZSUFGRkZODdd9/F1q1b8ccffyAwMBAAMHLkSBw+fBhPPvkkoqKikJmZiY0bNyI9Pd30fNCgQWjWrBleeOEFBAYG4syZM/j2228bXEciqoFERGQFS5YskQBIu3fvlj744APJz89PKioqkiRJku6//37p9ttvlyRJklq1aiUNHTrU9Lo1a9ZIAKTXX39ddr5Ro0ZJKpVKOnXqlCRJkrR//34JgDR58mRZuQcffFACIM2cOdO077HHHpPCw8Ol7OxsWdkxY8ZIAQEBpnqdPn1aAiAtWbKkxmvbtGmTBEDatGmTJEmSVFpaKoWEhEhdunSRrl27Ziq3du1aCYA0Y8YMSZIk6erVqxIA6c0336z23KtXrzbdNyKyH3aBEZHVPfDAA7h27RrWrl2L/Px8rF27ttrur59++gkajQb/+te/ZPuffvppSJKEdevWmcoBuK5c5dYcSZLwzTffYNiwYZAkCdnZ2aZHYmIicnNzsW/fvgZd3549e5CZmYnJkyfD09PTtH/o0KGIjo7Gjz/+CADw8vKCh4cH0tLScPXq1SrPZWwpWrt2LcrKyhpULyKqPQZARGR1zZo1Q0JCAlasWIFvv/0Wer0eo0aNqrLs2bNnERERAT8/P9n+jh07mo4bv6rVarRt21ZWrkOHDrLnWVlZyMnJwUcffYRmzZrJHklJSQCAzMzMBl2fsU6V3xsAoqOjTce1Wi3mzJmDdevWITQ0FLfddhvmzp0LnU5nKt+/f3+MHDkSr7zyCoKDgzF8+HAsWbIEJSUlDaojEdWMOUBEZBMPPvggHn/8ceh0OgwZMsTU0mFrBoMBAPDQQw9hwoQJVZbp1q2bXeoCiBaqYcOGYc2aNdiwYQNefvllpKSk4Ndff0WPHj2gUqnw9ddfY8eOHfjhhx+wYcMGPProo3j77bexY8cO+Pr62q2uRK6ELUBEZBP33nsv1Go1duzYUW33FwC0atUKFy9eRH5+vmz/sWPHTMeNXw0GA/766y9ZuePHj8ueG0eI6fV6JCQkVPkICQlp0LUZ61T5vY37jMeN2rZti6effho///wzDh06hNLSUrz99tuyMjfffDPeeOMN7NmzB8uXL8fhw4fx5ZdfNqieRFQ9BkBEZBO+vr5YuHAhZs2ahWHDhlVb7q677oJer8cHH3wg2//OO+9ApVKZRpIZv1YeRTZ//nzZc41Gg5EjR+Kbb77BoUOHrnu/rKys+lyOTGxsLEJCQrBo0SJZV9W6detw9OhR08i0oqIiFBcXy17btm1b+Pn5mV539erV64b7x8TEAAC7wYhsiF1gRGQz1XVBWRo2bBhuv/12vPjiizhz5gy6d++On3/+Gd999x2mTZtmyvmJiYnB2LFj8eGHHyI3Nxd9+/ZFamoqTp06dd05Z8+ejU2bNiEuLg6PP/44OnXqhCtXrmDfvn345ZdfcOXKlQZdl7u7O+bMmYOkpCT0798fY8eONQ2Dj4qKwlNPPQUAOHHiBAYOHIgHHngAnTp1gpubG1avXo2MjAyMGTMGALBs2TJ8+OGHuPfee9G2bVvk5+fj448/hr+/P+66664G1ZOIqscAiIgUpVar8f3332PGjBlYuXIllixZgqioKLz55pt4+umnZWUXL16MZs2aYfny5VizZg3uuOMO/Pjjj4iMjJSVCw0Nxa5du/Dqq6/i22+/xYcffoimTZuic+fOmDNnjlXq/cgjj8Db2xuzZ8/G888/Dx8fH9x7772YM2eOKd8pMjISY8eORWpqKj777DO4ubkhOjoaX331FUaOHAlAJEHv2rULX375JTIyMhAQEIA+ffpg+fLlaN26tVXqSkTXU0mV216JiIiInBxzgIiIiMjlMAAiIiIil8MAiIiIiFwOAyAiIiJyOQyAiIiIyOUwACIiIiKXw3mAqmAwGHDx4kX4+flBpVIpXR0iIiKqBUmSkJ+fj4iICKjVNbfxMACqwsWLF6+bWI2IiIgah3PnzqFFixY1lmEAVAU/Pz8A4gb6+/srXBsiIiKqjby8PERGRpo+x2vCAKgKxm4vf39/BkBERESNTG3SV5gETURERC6HARARERG5HAZARERE5HKYA0RERGQnBoMBpaWlSlej0XJ3d4dGo7HKuRgAERER2UFpaSlOnz4Ng8GgdFUatcDAQISFhTV4nj4GQERERDYmSRIuXboEjUaDyMjIG07SR9eTJAlFRUXIzMwEAISHhzfofAyAiIiIbKy8vBxFRUWIiIiAt7e30tVptLy8vAAAmZmZCAkJaVB3mEOEoAsWLEBUVBQ8PT0RFxeHXbt2VVv222+/RWxsLAIDA+Hj44OYmBh89tlnsjKPPPIIVCqV7DF48GBbXwYREVGV9Ho9AMDDw0PhmjR+xgCyrKysQedRvAVo5cqVSE5OxqJFixAXF4f58+cjMTERx48fR0hIyHXlmzRpghdffBHR0dHw8PDA2rVrkZSUhJCQECQmJprKDR48GEuWLDE912q1drkeIiKi6nB9yYaz1j1UvAVo3rx5ePzxx5GUlIROnTph0aJF8Pb2xuLFi6ssP2DAANx7773o2LEj2rZti6lTp6Jbt27YsmWLrJxWq0VYWJjpERQUZI/LISIiokZA0QCotLQUe/fuRUJCgmmfWq1GQkICtm/ffsPXS5KE1NRUHD9+HLfddpvsWFpaGkJCQtChQwdMmjQJly9frvY8JSUlyMvLkz2IiIjI+qKiojB//nylq6FsAJSdnQ29Xo/Q0FDZ/tDQUOh0umpfl5ubC19fX3h4eGDo0KF4//33ceedd5qODx48GJ9++ilSU1MxZ84cbN68GUOGDDH1wVaWkpKCgIAA04MrwRMRkaurnEtb+TFr1qx6nXf37t2YOHGidStbD4rnANWHn58f9u/fj4KCAqSmpiI5ORlt2rTBgAEDAABjxowxle3atSu6deuGtm3bIi0tDQMHDrzufNOnT0dycrLpuXE12UZDkoD0HUD2CaDzvYAnF3AlIqKGuXTpkml75cqVmDFjBo4fP27a5+vra9qWJAl6vR5ubjcOK5o1a2bditaToi1AwcHB0Gg0yMjIkO3PyMhAWFhYta9Tq9Vo164dYmJi8PTTT2PUqFFISUmptnybNm0QHByMU6dOVXlcq9WaVn5vVCvAF10Btn8ILIgDlgwGfvgXsP4FpWtFREROwDKPNiAgACqVyvT82LFj8PPzw7p169CrVy9otVps2bIFf/31F4YPH47Q0FD4+vqid+/e+OWXX2TnrdwFplKp8Mknn+Dee++Ft7c32rdvj++//97m16doAOTh4YFevXohNTXVtM9gMCA1NRXx8fG1Po/BYEBJSUm1x8+fP4/Lly83eNIkxRRkAUe+A/YsBn57E1j/b2DlQ8Db0cCG6UC2OSLH0bWAvly5uhIRkct44YUXMHv2bBw9ehTdunVDQUEB7rrrLqSmpuKPP/7A4MGDMWzYMKSnp9d4nldeeQUPPPAADhw4gLvuugvjxo3DlStXbFp3xbvAkpOTMWHCBMTGxqJPnz6YP38+CgsLkZSUBAAYP348mjdvbmrhSUlJQWxsLNq2bYuSkhL89NNP+Oyzz7Bw4UIAQEFBAV555RWMHDkSYWFh+Ouvv/Dcc8+hXbt2smHyjca1q8CiW4CCjJrLaf2BkjygJBe4sAdoebN96kdERPUy7P0tyMqv/p93W2nmp8UPT/azyrleffVVWQ5ukyZN0L17d9Pz1157DatXr8b333+PKVOmVHueRx55BGPHjgUA/Oc//8F7772HXbt22XQOP8UDoNGjRyMrKwszZsyATqdDTEwM1q9fb0qMTk9Pl00ZXlhYiMmTJ+P8+fPw8vJCdHQ0Pv/8c4wePRoAoNFocODAASxbtgw5OTmIiIjAoEGD8NprrzXOuYD+WF598OMVBHR/EOg1ATi/B/husth/6hcGQEREDi4rvwS6vGKlq9EgsbGxsucFBQWYNWsWfvzxR1y6dAnl5eW4du3aDVuAunXrZtr28fGBv7+/ackLW1E8AAKAKVOmVBsZpqWlyZ6//vrreP3116s9l5eXFzZs2GDN6inHYAB2f2x+PngOEBgJeDUBvJsAQa0Bt4pZRT0DzOVO/QLc8ZJ960pERHXSzE+Zf8qt+b4+Pj6y58888ww2btyIt956C+3atYOXlxdGjRqF0tLSGs/j7u4ue65SqWy+aKxDBEBUjVO/AFfPiO02twM3/7P6sn5hQGgXIOMQcHE/UJgN+ATbo5ZERFQP1uqGciRbt27FI488gnvvvReAaBE6c+aMspWqhuIzQVMNdn1k3u5TizkT2hmH+EvAX5tsUiUiIqLqtG/fHt9++y3279+PP//8Ew8++KDNW3LqiwGQo7r8F3Bqo9gOaAncVIsE7nbmGbVx6pfqyxEREdnAvHnzEBQUhL59+2LYsGFITExEz549la5WlVSSJElKV8LR5OXlISAgALm5ucrNCbT+38COBWI74RWg37Qbv6a8FJgTBZQVAj7NgKdPAGrGuERESisuLsbp06fRunVreHp6Kl2dRq2me1mXz29+Ojqi0kLgj8/Ftpsn0HN87V7n5gG06S+2C7MA3QHb1I+IiKiRYwDkiA58JebzAYAuo8SIr9pqZ7HUB7vBiIiIqsQAyNFIErDLYuh7n8fr9vq2lgFQavXliIiIXBgDIEdzdhuQeVhst+gNRMTU7fVNWgNN2ort87uA4lyrVo+IiMgZMAByNMbcH6B2Q9+rYhwNZigHTv/W8DoRERE5GQZAjib7hHm74z31OweHwxMREdWIAZCjyb8kvvo0A9zrOVQy6hZAUzHV+alUkVdEREREJgyAHIlBD+TrxLZfeP3P4+EDtOortnPPAZlHGl43IiIiJ8IAyJEUZAKSXmz7RzTsXJYzR299r2HnIiIicjIMgBxJ/kXzdkNagAAgZhzgFSS2D34FZJ9s2PmIiIjqaMCAAZg2bZrS1agSAyBHknfJvO3fvGHn8vQH4qeIbckAbJ7bsPMREZFLGTZsGAYPHlzlsd9//x0qlQoHDjTeFQcYADmSfMsAqIEtQAAQ9w/Aq2IW6UNfA1knai5PRERU4bHHHsPGjRtx/vz5644tWbIEsbGx6NatmwI1sw4GQI4kz4pdYACg9QNu+ZfYlgzA5jkNPycREbmEu+++G82aNcPSpUtl+wsKCrBq1SqMGDECY8eORfPmzeHt7Y2uXbviiy++UKay9cAAyJHIWoAamARt1PtxwLup2D70DZB5zDrnJSIip+bm5obx48dj6dKlkCymU1m1ahX0ej0eeugh9OrVCz/++CMOHTqEiRMn4uGHH8auXbsUrHXtuSldAbJg7RYgAND6ArdMBTbOACABm2cD9y+1zrmJiKj+/ttfjP61N98Q4B+ba1X00UcfxZtvvonNmzdjwIABAET318iRI9GqVSs888wzprJPPvkkNmzYgK+++gp9+vSxRc2tigGQIzG2ALl7A54B1jtv7/8Dtr0PFGYBh9cAtx0BQjtZ7/xERFR3BZny0b8OKDo6Gn379sXixYsxYMAAnDp1Cr///jteffVV6PV6/Oc//8FXX32FCxcuoLS0FCUlJfD29la62rXCAMiRGEeB+YUDKpX1zuvhI1qBfn4JgAT8NpetQERESvMNaRTv+9hjj+HJJ5/EggULsGTJErRt2xb9+/fHnDlz8O6772L+/Pno2rUrfHx8MG3aNJSWltqo4tbFAMhRFOcBpfli21r5P5ZiHwO2vitagY58L4Ita4w0IyKi+qllN5TSHnjgAUydOhUrVqzAp59+ikmTJkGlUmHr1q0YPnw4HnroIQCAwWDAiRMn0KlT4+hhYBK0o7BMgLZW/o8lD2+g5wSxLenlq84TERFVw9fXF6NHj8b06dNx6dIlPPLIIwCA9u3bY+PGjdi2bRuOHj2Kf/zjH8jIyFC2snXAAMhRWCZA26plptcEABVda/uWibXHiIiIbuCxxx7D1atXkZiYiIgI0Uvx0ksvoWfPnkhMTMSAAQMQFhaGESNGKFvROmAXmKOQtQDZoAsMAAJbAu0SgFMbxSKpp1KBmwbZ5r2IiMhpxMfHy4bCA0CTJk2wZs2aGl+XlpZmu0o1EFuAHIWsBchGARAAxCaZt/cusd37EBEROTAGQI7CFpMgVqV9ormF6cR6IPeC7d6LiIjIQTEAchR5Nk6CNtK4AT0fFtuSAfjjM9u9FxERkYNiAOQojJNhqdSAb6ht36vnePE+ALDvU0Bfbtv3IyIicjAMgByFsQXIJ0S00thSQAvRFQYAeRdEUjQREdlc5URiqjtr3UMGQI5AXwYUVMydYK/JCS2Tofcsts97EhG5KI1GAwCNZpZkR1ZUVAQAcHd3b9B5OAzeERRkAKiIaG01BL6ydglAQKQYDn9yI5CTLobJExGR1bm5ucHb2xtZWVlwd3eHWs32h7qSJAlFRUXIzMxEYGCgKaisLwZAjsAyAdpeLUBqjcgF2vQGAEksj9F3in3em4jIxahUKoSHh+P06dM4e/as0tVp1AIDAxEWFtbg8zAAcgSWqwHbcgRYZdFDKwIgAGe3MQAiIrIhDw8PtG/fnt1gDeDu7t7glh8jBkCOIM9OcwBV1qwj4BUEXLsKpG8DDAaAzbJERDajVqvh6empdDUITIJ2DPl2mgW6MrUaaNlXbF+7CmQesd97ExERKYgBkCPIs8M6YNWJusW8fXarfd+biIhIIQ4RAC1YsABRUVHw9PREXFwcdu3aVW3Zb7/9FrGxsQgMDISPjw9iYmLw2Wfy2YwlScKMGTMQHh4OLy8vJCQk4OTJk7a+jPrLVyAJ2qiVRQB0Zot935uIiEghigdAK1euRHJyMmbOnIl9+/ahe/fuSExMRGZmZpXlmzRpghdffBHbt2/HgQMHkJSUhKSkJGzYsMFUZu7cuXjvvfewaNEi7Ny5Ez4+PkhMTERxcbG9LqtujAuhevgBWj/7vndYV0AbILbPbgM4SRcREbkAlaTwtJRxcXHo3bs3PvjgAwCAwWBAZGQknnzySbzwwgu1OkfPnj0xdOhQvPbaa5AkCREREXj66afxzDPPAAByc3MRGhqKpUuXYsyYMTc8X15eHgICApCbmwt/f//6X1xtSBLwnwigrAgIvgmYstu271eV5Q8AJysCyMk7gZBo+9eBiIiogery+a1oC1BpaSn27t2LhIQE0z61Wo2EhARs3779hq+XJAmpqak4fvw4brvtNgDA6dOnodPpZOcMCAhAXFxcrc5pd8W5IvgB7DsE3hLzgIiIyMUoOgw+Ozsber0eoaHyxT9DQ0Nx7Nixal+Xm5uL5s2bo6SkBBqNBh9++CHuvPNOAIBOpzOdo/I5jccqKykpQUlJiel5Xl5eva6nXvIUGgFmqVWlAKj3Y8rUg4iIyE4a5TxAfn5+2L9/PwoKCpCamork5GS0adMGAwYMqNf5UlJS8Morr1i3krWl1CSIlsK7A+4+QFkhcGar6JZTqZSpCxERkR0o2gUWHBwMjUaDjIwM2f6MjIwap7lWq9Vo164dYmJi8PTTT2PUqFFISUkBANPr6nLO6dOnIzc31/Q4d+5cQy6rbpSaBNGSxh1oGSe2C3TAlb+VqQcREZGdKBoAeXh4oFevXkhNTTXtMxgMSE1NRXx8fK3PYzAYTF1YrVu3RlhYmOyceXl52LlzZ7Xn1Gq18Pf3lz3sxnIIvFItQACHwxMRkUtRvAssOTkZEyZMQGxsLPr06YP58+ejsLAQSUlJAIDx48ejefPmphaelJQUxMbGom3btigpKcFPP/2Ezz77DAsXLgQgFpybNm0aXn/9dbRv3x6tW7fGyy+/jIiICIwYMUKpy6yeLAdIwQAoqp95++xWoNcE5epCRERkY4oHQKNHj0ZWVhZmzJgBnU6HmJgYrF+/3pTEnJ6eDrXF+lSFhYWYPHkyzp8/Dy8vL0RHR+Pzzz/H6NGjTWWee+45FBYWYuLEicjJyUG/fv2wfv16x1x/RTYJYnPl6hHRE3DzAsqvMQ+IiIicnuLzADkiu84DtOhWQHcAUGmAl7MAtXVWua2XZcOA07+J7akHgKBWytWFiIiojhrNPEAEcwuQX5iywQ9w/XB4IiIiJ8UASEnlpUBhlthWMgHaSJYIzQCIiIicFwMgJRVYTMyoZAK0UYtYQOMhtk9vBgx6ZetDRERkIwyAlGQ5B5CfQnMAWXL3MrcC5Z4DDq9Wtj5EREQ2wgBIScW55m3vJsrVw9ItU83bv70JGAzK1YWIiMhGGAApqazQvO3urVw9LLUZAERWzAqddQw4skbJ2hAREdkEAyAllV0zb7t7KVcPSyoV0P958/PNc9kKRERETocBkJJKLVqAPHyUq0dlbe8AWvQW21lHgaPfKVsfIiIiK2MApCRZC5CDdIEBFa1AL5ifsxWIiIicDAMgJZUVmbcdKQACgHYDgea9xHbmEeDYD8rWh4iIyIoYACnJMgDycLAASKUCBkw3P2crEBEROREGQEoqtWwBcpAkaEvtEsQiqQCQcQjY9q5YJJWIiKiRYwCkJFkXmAMlQRtVbgX6ZRbw0zOAvlyxKhEREVkDAyAllTl4CxAA3DRInhC9+xPgi9FAcZ5ydSIiImogBkBKsuwCc6Rh8JXdPh0YsQhQu4vnp34BFicCOenK1ouIiKieGAApyZFHgVUWMxYY/x3gFSSeZx4Blt4tH8pPRETUSDAAUpJlAOTmqVw9aivqFuD/UoEmbcXznLPA3qWKVomIiKg+GAApydh64u4NqBvJt6JpW+CBT83Pt7wj78ojIiJqBBrJp66TMi6F4agJ0NUJ6wJ0Gi62CzKAPYuVrQ8REVEdMQBSkqkFyIEToKvT/wUAKrG9db58XTMiIiIHxwBIScYcoMbWAgQAoZ2AziPEdmGWGB5PRETUSDAAUookmVtNHG0ZjNqStQK9C5QUKFodIiKi2mIApJTyEgAVy0o0xi4wAAiJBrqMFNtFl4FdHylbHyIiolpiAKSUxjALdG30fx5QVfwYbXsPKMisulxpIfD3ZuDYT0BJvv3qR0REVAU3pSvgshx5Jfi6aHYT0PV+4MBK4NpV4K32QFBrIKIHEN4dyL8EpO8AdAcBSS9e4xMC3PES0OMhQK1Rtv5EROSSGAAppbQRzQJ9I/2fBw6vBvSl4vnV0+Jx+NuqyxdmAj/8C9j1MTD4P0Dr2+xXVyIiIrALTDmNaRmMG2naFpjwA9DnH0CLPlXMaq0CQjoBvZKAjsPMuzMOAsuGAaseAfRl9qwxERG5OLYAKcVZcoCMWt4sHgCgLweyjgEZhwHvJkCL3oBXoLnsma3AhunApT/F88Orge5jgZsS7V5tIiJyTWwBUkpjWQm+PjRuYrbo7qOB9nfKgx9ArCn2eBpw+0vmfWe32bOGRETk4hgAKcWZusDqQ60Geo43Pz+3U7m6EBGRy2EApBRXD4AAwC8UCIoS2xf2VcyNREREZHsMgJTiLMPgGyrSmDdUAlw6oGxdiIjIZTAAUkqpkyVB11fLOPP2uR3K1YOIiFwKAyClGFeCBxrvUhjWYGwBAsSEiURERHbAAEgpZYXmbVduAWoWDWgDxPa5nWKRWGvJvSCG5BMREVXCAEgpzjwMvi7UaiCyt9guzAKu/G2d86a+CrzTCVh2N2AwWOecRETkNBgAKYWjwMwsu8GsMRz+8Grg97fFdvp2IONQw89JREROhQGQUpxtJuiGkCVCNzAAyj4FfPekfN/pzQ07JxEROR2HCIAWLFiAqKgoeHp6Ii4uDrt27aq27Mcff4xbb70VQUFBCAoKQkJCwnXlH3nkEahUKtlj8ODBtr6MurFMgnblLjAAaN4LUFWsCp/egACotAj4ajxQmi/f/3da/c9JREROSfEAaOXKlUhOTsbMmTOxb98+dO/eHYmJicjMzKyyfFpaGsaOHYtNmzZh+/btiIyMxKBBg3DhwgVZucGDB+PSpUumxxdffGGPy6m9UiZBm3j4AOHdxHbWUeDa1fqd56dngczDYrtZNOAbKrbPbgPKSxteTyIichqKB0Dz5s3D448/jqSkJHTq1AmLFi2Ct7c3Fi9eXGX55cuXY/LkyYiJiUF0dDQ++eQTGAwGpKamyspptVqEhYWZHkFBQfa4nNqTDYN38RwgAIi07AbbXffX7/sM2P+52Hb3AR74FGhzu3heVgScr8c5iYjIaSkaAJWWlmLv3r1ISEgw7VOr1UhISMD27dtrdY6ioiKUlZWhSZMmsv1paWkICQlBhw4dMGnSJFy+fLnac5SUlCAvL0/2sDljDpBGC6g1tn8/RxfZgAkRz24DfnrG/HzYu0CzDkCbAeZ97AYjIiILigZA2dnZ0Ov1CA0Nle0PDQ2FTqer1Tmef/55REREyIKowYMH49NPP0VqairmzJmDzZs3Y8iQIdDr9VWeIyUlBQEBAaZHZGRk/S+qtowBkCsvg2GppeWEiHXIAzq/F1j+AFBeLJ7HPgp0u19st+lvLscAiIiILLgpXYGGmD17Nr788kukpaXB09PTtH/MmDGm7a5du6Jbt25o27Yt0tLSMHDgwOvOM336dCQnJ5ue5+Xl2T4IMs4D5MqzQFvyjwACWgK56cCFvYC+DNC41/wa3UHg83vNSc/tEoDBs+XnDL4JyD4hzlmcB3j6m49LEnB+DxDYUizMSkRELkPRFqDg4GBoNBpkZGTI9mdkZCAsLKzG17711luYPXs2fv75Z3Tr1q3Gsm3atEFwcDBOnTpV5XGtVgt/f3/Zw+aMLUCungBtyTgcvvwaoLvBwqhZx4FPRwDFueJ51K3A6M8BN628nLEbTNIDZ7fKj219F/hfArAwHiipNHKMiIicmqIBkIeHB3r16iVLYDYmNMfHx1f7urlz5+K1117D+vXrERsbe8P3OX/+PC5fvozw8HCr1Nsq2AV2Pcs8oJq6wbJOAMvuAYqyxfMWfYCxX1YdTFaXB5SvAzbPFdtFl0VLEBERuQzFR4ElJyfj448/xrJly3D06FFMmjQJhYWFSEpKAgCMHz8e06dPN5WfM2cOXn75ZSxevBhRUVHQ6XTQ6XQoKCgAABQUFODZZ5/Fjh07cObMGaSmpmL48OFo164dEhMTFbnG65SXAoaKNao4AszMMgD6cwVwYZ/8+LWrwIYXgYV9gYKKHLHw7sBDXwNa36rPGdUPUFX8mP9tMSFi2mz5emycLZqIyKUongM0evRoZGVlYcaMGdDpdIiJicH69etNidHp6elQq81x2sKFC1FaWopRo0bJzjNz5kzMmjULGo0GBw4cwLJly5CTk4OIiAgMGjQIr732GrTaSt0jSuEyGFUL7Qx4BohuLd1B4OPbgVa3APFTgNxzQFqKfI6gkE7AQ6vFa6rjGQBE9AQu7BFzDOXrRC7Qvk/l5XQHbXNNRETkkFSSZM3lt51DXl4eAgICkJuba5t8oLyLwLyOYjv6bmDMcuu/R2N18Gtg3fPm7q2quHmKoKjfNEDrd+Nzpr4G/P6W2L73I+Do98CxtfIyIZ2BydvqXW0iIlJeXT6/Fe8Cc0lcBqN6XUcBTx0Sc/kE31TF8QeAKXuAgS/XLvgB5HlA2943Bz++YUCTtmI7+zhQXtKgqhMRUeOheBeYS5Itg8EusOu4ewG9HgF6jAdO/gzs+Z/I4+n/nFg3rK4i+wBuXmJ0WYZFV9ft/xYjw678JXKyso6JnCIiInJ6DICUwByg2lGrgQ6DxaMh3LRAq3jgr1/N+4I7ADHjKobRrxT7dIcYABERuQh2gSnBMgDiMHj7sOwGA4A7XwE0bkBYV/M+jgQjInIZDICUUGrZAsSJEO2ircUM4K1uAW6qaFWyDIA4EkxZJfnAqkeAVUnA5b+Urg0ROTl2gSlBthI8k6DtIqwLMOh1sSTGoNcBlUrs9wkWydAFOhEASZL5mK2Ul4rV6f9KBS4dADoOA3pNsO17NgZHvgcOrxbbx38CBs4A4v7JxYKJyCYYACnBcgI+doHZT98nq94f1hU4pQOKc4C8C0BAi5rPI0nioa5DA6okAYe/FcP8T/8GlBaYj53aKOrQvGftz+eM8i6at8uLgQ3/Bo58Bwz/EAhup1y9iMgpsQtMCbIWIAZAigvrYt6uTTfYj08DrwYBi4eIVguDvuby5aXA908CXz8qWjYsgx+jDS+KIMmVFedcv+/cTmDRLcDhNfauDRE5OQZASuAweMcSahkA3SAR+u80MSwfANK3AV89DLzXA9j+oZhhurKiK8Bn9wJ/fGbe5x0s5jMascg8D1H6NjFBoyu7lmPeHvo2ENRabJcXAz9MBfRlilSLiJwTu8CUIGsBYhK04mQjwWpoATIYgJ9fvn5/zllgw3Qg9RWgze1AhyEiybo4F1jxAHD1tCin0QL3vA90vd/cfebpD3z5oNjeOEO8rvKK9q7CsgWow11A97HAF2OB05vFsfQdQOtblaodETkZtgApQTYMnknQimvaTkyUCNTcAnTwK0B3QGyHdQPGfSMfXVZeDJxYB/zwL+Dtm4BF/czBj08IkPQT0H20PHeow11AVMWH+tUzwM7/3ri+ZcXyfBlnYdkC5Bkofjd6PGTed3KDvWtERE6MAZASZF1gbAFSnFoDhFSszXblb6CkihydsmtiTTGjQa8B7ROAh78FJu8EYh8DfEPlr9FXLK0R2hV4/FegRez151WpgMT/AKgYefbbm0BhNeugXT0D/PySCK7mdQQ2z63LVTo+YwuQxsP8e9EuQcwCDgAnflakWkTknBgAKYFJ0I7HlAgtAZlHrj++cxGQd15st7tTPrFiSDRw9zwg+ZgIdG571pxX1Gk48Oh6IDCy+vcO7wb0GCe2S/LEqvcAoC8Hcs8DJzcCXzwIvBsj1jIrzhXHN/0HOP17PS/YARlbgDwDzVMReDcBWvQR29nHgSunlagZETkh5gApgV1gjiesm3lbd1CsH2ZUeBn4fZ7YVqmBO1+t+hxqtVirrHkv4I6XxOKqtc3nueNl4NBqMUXCnsXAiQ2im0uqYoSZSg1IBgASsPofwKStgFdQ7d7HkRlbgLwC5ftvGgSc2yG2T/4MxP3DnrUiIifFFiAllHEmaIcTWsNQ+N/mipYZQKwfFtqpduesSzKzXxhw61NiWzIAueeuD378IkSg9PRxc95Q3gVgbXLNQ+gNBtHtWngZuLgfOPCV6M5b+RDwSQKwf0Xt62kr+jLz9ACVg7n2iebtE8wDIiLrYAuQEkq5GKrDCe1s3rZcEyzzKLD7E7Ht7g3c/qLt6hA/BTj5i2jt8G4qJmQMiBRfW8YD0UMBjbsoe+8iYGFf0R12+Fsxeqz7aHOdt8wXEyyWFJhzkapzfo8IOjoMsd213UjlBGhLoZ0B/+Yi2DuzRQRzbDklogZiAKQEYwuQ2t38gUbK8vQHAluJIe0ZR8TkhrqDwOcjAUO5KBM/BfAPt10d3L2AxzaI1pAb/VwEtADung98nSSe//QMoPUD9i8Hjq2t4xtLwDf/Bzy6QT4ppD1ZDoGv3AWmUgHtBwF7l4hg7u/NQPRd9qwdETkhBkBKMAZAXAbDsYR1FQFQWSGwbxnw8wygNF8cC+0C3DLVPvWobVDc5T7RJXTgS9FF9+VY+XFtABDYEnD3BNw8RYDl0wwIvglo1kF8/fV10YJUWgB8MUYkcfuGWP+abqSmFiBAtHDtXSK2T25gAEREDcYASAnGLjB2fzmWsK7m1pO1T5n3t4wHxn4BaH2VqVdN7npTzCKdk27e5xsm1j3r9ciN6zziQzG8/uI+kXe08iFg/PciaLKnmlqAAKD1bSKIKy8Ww+HtsWgtETk1JkEroYwBkEMKraL756YhwMOrHXeUlac/cP9SwL+FaNG5+x1g6p9A3ym1C9jcvYAxK0SCNSDW3vphqv3XJbtRC5CHtznxO/9i1Wu2SZJI9E7fKRK7D6++8TptROSy2AKkBHaBOSbLJTEAMeJr2HuAxsF/TZr3ApIP1//1/uGihWvxYKD8muhSC+1kvy4/4MYtQABwU6JI7AZEN1h4NzHC7eBXwJ4lQNax6xdUHfS6aA0jIqqELUD2pi8H9KVimy1AjiWwJRB9t0hOv/VpYPgCxw9+rCUiBrjPYhmOX2aJhV/t5UYtQIBIhDY68TNwKhX4721iLqRzO6peTf7g11asJBE5EwZA9lbGIfAOS6UCxiwHpp8DBs5wvRyTTsOB254T25IBWJUkzy2ypdq0AAW1AppVLFlyfhfw+X3yxWsDIsUM3bGPiRF9AHBpP1CQaf36ElGjxwDI3rgMhuNz5ckpB0w3t7RcuyKSoi1/Zm2lNi1AgJgVurLw7iJx+6lDwPjvxLIkXUaaj59KtVYtq1dVzlRZsWhF++UV4KPbgf+0AD67D8g6bvv6kFzeRTGHFHPCyIKLtO87kDKLhVCZA0SORq0G7vtIfGBfPQ1c+lPMND3iQ9u2iNWmBQgAOt8LbH1XbAe2BAbOBDrfJ+ptqf2dwJaK5UtO/QLEVJoiwFou/gH89lbFDNWS+KfG3VsE0fmXxKg1S3+lAgtvEUnqtz1n/htQWgSc3gykbxejDpWclNKZXLsqvj87/wsYyoDbXwL6P6t0rchBMACyN1kLkAu3NJDj8goSXYGfJIgu2z9XiIkX+04BPANs8561bQGK6AE8vAYougx0HFb9ciMtegNafzE/0l+p4j9/tcZ69U3fCfz2pjkp26gkz7xsSmXG+hjKgC3vAAe/AXpNECPvTv9mDpa2fyiWO/Fpar36Oqu/NokAJ+es+NlofRvQuj8QFAXs+R+weY4IgowOfc0AiEwYANmbbBkMTudPDiq0M3DP+8A3j4nnv80Ftr0nksRjHhS5NtYMKIwtQBrtjf8xaHv7jc+ncRd1PPq9+AC8sA+I7H3j1+XrgJJ8sRSJZ4D5GnMvABf2isfZrcD53fLX+YSI0XRl18TveFmRmJk76lZRj9a3iSkLfp8HbJ0vBkLkpgO/vnZ9HQxlIrepzYDrj0kSkL4D8AkGgtvf+HoaI4MByD4BXNgjvm+SHoi8GYjqBwRGijIX9gGpr8gT9XPPie83YJ4zqrKsY+J77Bdm88sgx8cAyN7YBUaNRddR4gPjtzfF8/Ji8R/0oa8Bv3DRHdX5PqBFbMO7x4z/pXsFWq+rrV2C+QPx1C83DoCOfA+smiASwAEAKtEaptYAhVlVvyagJdBvKhDzUO0mj7zjRaDbaOCnp+Uf3n4RopXt/C7xPPtk1QHQwa+Bb/9P1K33YyJZ31atcg2Rd1HcM41HxZI/bqJr0KfZ9d9fSRLr753cKLoBL+y7vhVt71LxNShKJLuf+V1+XO0uAkcjy+Cn22hxfP/n4vnp34Fu91vjKqmRYwBkb+wCo8bkjpdEq8/+FWK+HWOgkn8J2PGheAREAp1HALGPAk3a1O99jF1gNXV/1VW7BPP2qY3A7dOrL5ufAfzwL4vgBwAkkQheleCbxDxJ3UbXfT2/4HaiG+/EeuDyKdFlE9ZV5P8sqcj9qS5R+uTP5rrt/gQ4+gMwZA7QaUTtAkd9mZhE8vxusQiuoQyIjANa9RUTgTa0Ve/K30Dqq2ISyqp4+AJN24n7F9xedF2dShU/T7Vx9Yx4GAVFibyeziNEEHX6N/HQHRKtmHe8BDTvKYJNUwCUxgCIADAAsr9SixYgdoFRYxARIx6DXhcTEO5fIf5bN/7HnXsO2PY+sO9T4KnDouunLvRl5pbRmhKg6yqgORDSGcg8LFoVCrNF11FlkiSWPjEGd2FdxXIiRZfFo+waENJRTDjZIlZ8bWgXikp1faJzcAfzdvaJql+XeVT+vCADWPUI0HagyIExX5ToZistFF1ypQXi+i/9KSa7tGQMVrQBQMs4EcR6NzU/fENF65RfePXzYhVdATbPFUGZZUtMZaUFYmqCS/urL+MbKnK4WsQCzWPFvrNbRcvN+V3iunxCgP7PAT0nAG4eokxED/GoagLPyDjRvaovEQESERgA2R9bgKixcvMQiccdh4lg4diPwKFvxX/Xkh4ozhWjolrfVrfz1jYBuj7aDRQBECSRMFvVf/4HvgKO/yi2fZoBD3+nTAKyT1PAq4lodaoqANKXA9kVLUMBLUULx4l14vlfqeLRECW5Fi1MVVBpAP8I0eLn4Q1AVdHqpBJ5SSW55rLewWLBWskg6m0oA4rzRItXzll5S5ubp8iVan+naLVr0ub61qzWtwIDXhB/P3POmRf5rS13LyCyj+g6y0kHrpwGmrSu/evJKTEAsjfLiRA92AJEjZRXENDjIfHYs9i8eGzm0boHQLUdAl8f7e8UyduA6AarHADlXQLWPWd+fvc7yo6+atZBdIXlXxIBg6e/+diVv82zyDfvKdaAO7YW+Ok5sT5abQS2BFr0EcFAi96A2k2839mtwNlt1ec6ASLIzT0nHtVx8wLinxCtMJZ1t1RWLK7l8knRWtgyvvb/DLp7Ac1uql3Zytr0N+cOnf6NARAxALI72UzQbAEiJ2C5iGxGPdYks2ULUOTNIu+ktEDkmhgM5jmDJAlYO80cgHUZJVq3lBTcXgQkgEiEbtHLfCzziHk7pJNoJek4TLSaXPzDHBwZaTzEP1nuPuKr1rfq7snwbkDcP8T9yEkXM2cXZYvuv8Js0c2Wky4Cn5xzVedFqTRirqXbXxStRDVx9xRrzYV2qt09sZbW/QG8LrZPbxZTEJBLYwBkb7Jh8BwFRk4gpKN52/JDurZs2QLk5iE++I7/KD7UL+0XrScAsH+5SEQGRE7JXW9a973ro3IekCwAssj/sbzn7l4iibmhVCqx3EhQq5rLlRaJXBrj7NeSQfwtc/RRrRE9AQ8/oDRftABJkustd0MyXArD3tgFRs5G6ye6VgDxIW0w1Fy+Mlu2AAFAe4vRYCd/Bo6vB5bdA3z3hHn/sPmAdxPrv3ddBVt072RXGglWuQVIKR7eogvUu4l4+AQ7fvADiARuY6BYmHV9Qjm5HLYA2Ru7wMgZhXQW3SSlBWKCv6Co2r/Wli1AgHw4fFrK9ce7jwWih1r/fevDMr8l+6T8mPEDW6Nl/kp9tekvRjICohXI3t1w5FDYAmRvnAmanJHlB0lGHbvBbN0CFNhS3rVk1KQNMORN4J4PrP+e9RXQUoyKAuRzAZUVA1f+EtvNOlh3Fm5X0rq/efv0ZuXqQQ7BIQKgBQsWICoqCp6enoiLi8OuXbuqLfvxxx/j1ltvRVBQEIKCgpCQkHBdeUmSMGPGDISHh8PLywsJCQk4efJkNWe0M7YAkTOy7JLJrGMitK1bgAD56K/WtwFjvwSm7AXiJlY/t40S1GqgacUSF1dPizmSAJEPZBw6rmT3V2MX0knMbQSI1eH15crWhxSleAC0cuVKJCcnY+bMmdi3bx+6d++OxMREZGZmVlk+LS0NY8eOxaZNm7B9+3ZERkZi0KBBuHDhgqnM3Llz8d5772HRokXYuXMnfHx8kJiYiOLiKtaGsTdZDlAj6Dcnqo3QzubtOrcAWSxW6RVknfpU1i8ZeOgbYPJOYMIPYhLCyivIOwpjN5ihXAwXB6pPgKa6UavN0zSU5ImJIcllKf4XYN68eXj88ceRlJSETp06YdGiRfD29sbixYurLL98+XJMnjwZMTExiI6OxieffAKDwYDUVDEJmCRJmD9/Pl566SUMHz4c3bp1w6effoqLFy9izZo1dryyasgmQmQARE6iaTux3hJQ95Fgtu4CA0SXUbsEICTaNue3JlkidMWEiI6SAO0MatsNdi0H2LOk+mU9qNFTNAAqLS3F3r17kZBgTlJUq9VISEjA9u3ba3WOoqIilJWVoUkTMYLj9OnT0Ol0snMGBAQgLi6u1ue0KeNSGCqNmKeDyBlo3EVuCiCSd8tLav9ae3SBNSaWAZAxD4gtQNZjOVFnVQFQ7nlgw4vAO53FPFGrHgHSd9qrdmRHinZ+Z2dnQ6/XIzQ0VLY/NDQUx44dq9U5nn/+eURERJgCHp1OZzpH5XMaj1VWUlKCkhLzH+y8vLwqy1mFsQXIw4dzUJBzCekkFqSU9CIICuty49cA5hYgjZZ5cUClFqCK3EVjAOThJ9blovpr0kYs55F7Tsx+/cVYkRfk00wEP4e/Fd2Pls5uEeukkVNxoOy/ups9eza+/PJLpKWlwdOzDuvCVJKSkoJXXnnFijWrgTEHiH/oydmEdgIOVmxnHql9AGRsAWLrj9C0HQAVAEnMBVSSL6YWAETrD/9xahiVSnSD7f9czJ59/Keqy6ndzIGQ7mDVZahRU7QLLDg4GBqNBhkZGbL9GRkZCAurebXlt956C7Nnz8bPP/+Mbt26mfYbX1eXc06fPh25ubmmx7lzNax101DGLjDm/5CzscxNqcuSGMYWIFvl/zQ27p7m2ZizTwKZFq3h7P6yjr5PAk3aVn3MqwnQ/3ngqcPmKQkYADklRVuAPDw80KtXL6SmpmLEiBEAYEponjJlSrWvmzt3Lt544w1s2LABsbGxsmOtW7dGWFgYUlNTERMTA0B0ae3cuROTJk2q8nxarRZardYq13RDxi4wBkDkbGRD4WuZCK0vA8oq/ilgC5BZcAfg6hkxsaTlKu9MgLaOkGjgX/uA8lKx5llRtlj3zKAHWsWbZ+kP6QRc3Adc/gsoKRDrqZHTULwLLDk5GRMmTEBsbCz69OmD+fPno7CwEElJSQCA8ePHo3nz5khJETO4zpkzBzNmzMCKFSsQFRVlyuvx9fWFr68vVCoVpk2bhtdffx3t27dH69at8fLLLyMiIsIUZCnGYADKjTlADIDIyQS0ALQBQElu7YfC22MEWGMU3N48Y/HhNeb9bAGyLjcPwD9cPKoS1lUEQJBEUB/Zx67VI9tSPAAaPXo0srKyMGPGDOh0OsTExGD9+vWmJOb09HSoLebrWLhwIUpLSzFq1CjZeWbOnIlZs2YBAJ577jkUFhZi4sSJyMnJQb9+/bB+/foG5QlZRbnlEHjmAJGTUanEB/S5HUDeeRHc3KhVhyPAqtbMYubqLMsRYGwBsquwruZt3QEGQE5G8QAIAKZMmVJtl1daWprs+ZkzZ254PpVKhVdffRWvvvqqFWpnRVwGg5xdaCcRAAFi5FKr+JrLswWoapYjwYy8gwHfZvaviyuTBUCHlKsH2YTiEyG6FM4CTc6urktisAWoalUFQOz+sj/LGc6ZCO10GADZE9cBI2dX1yUx7LEMRmPk3US0+Fhi95f9af3EvEGAGNlo0CtbH7IqBkD2xC4wcnaWrRS1GQnGLrDqNau0gj1bgJRh7AYrvyZGg5HTYABkT2wBImfnFQT4NxfbmUcASaq5PLvAqle5G4wtQMqonAhNToMBkD3JcoDYAkROyvhBXZwL5F2suSxbgKp3XQDUCBZydUahlgEQ84CcCQMge2ILELmCunSDsQWoes0sAiD/FoBngHJ1cWVhDICcFQMge5LlAHEUGDkpWSL0DUaCsQWoes0sWnxC2f2lGP8IsTwGIBb7JadRrwDo3LlzOH/+vOn5rl27MG3aNHz00UdWq5hTYhcYuYK6rAnGFqDqBbQA4qeIrrB+yUrXxnWpVOZWoIIMID+j5vLUaNQrAHrwwQexadMmAIBOp8Odd96JXbt24cUXX3S8yQcdCbvAyBU062BeRNI4KWJ1jC1AGi1/J6qS+AYwZfeNJ5Qk27LsBstgN5izqFcAdOjQIfTpI6YE/+qrr9ClSxds27YNy5cvx9KlS61ZP+fCYfDkCty0QIveYjsnXTyqY2wBYusPObKwbuZt5gE5jXoFQGVlZabV03/55Rfcc889AIDo6GhcunTJerVzNmwBIlcRdat5+8zW6ssZW4CY/0OOLKyLeZsBkNOoVwDUuXNnLFq0CL///js2btyIwYMHAwAuXryIpk2bWrWCToVLYZCriLrFvH1mS9Vl9GVAWaHYZgsQObLgmwCNh9hmAOQ06hUAzZkzB//9738xYMAAjB07Ft27dwcAfP/996auMapCmeVq8AyAyIk1jxV5PQBw5veqy3AEGDUWGnfz9A6XT8nTGWzl8l/A7/OALe8AJzaIruQbTSxKdVKv1eAHDBiA7Oxs5OXlISjIvH7PxIkT4e3ND/ZqlRaatxkAkTNz9wQi+4jgJ+es+OMd2FJehuuAUWMS1hW49CcgGYDMo0CLXrV7XWmR+D0oyAACWwFN2wJ+EYC6ivaH8hLg2Fpgz5Kq/3Hw8BX1uO0ZoF1Cw66H6hcAXbt2DZIkmYKfs2fPYvXq1ejYsSMSExOtWkGnYtkCxGHw5Oyi+pn/iJ/ZCsRUCoA4BJ4aE1ki9IGaA6Dc86LV5sQG4PRmoLxYftzNEwhqLRZbVakrHiog6xhQdLn685YWAOnbgc9HAf2fA/o/D6g1DbsuF1avAGj48OG477778M9//hM5OTmIi4uDu7s7srOzMW/ePEyaNMna9XQOTIImVxLVz7x9ZgsQM1Z+nF1g1JhYDoW/9Of1x7NOAEe/B47+AFzaX/O5youBrKM1l2naDug5AfALFzOqZx4VEzHmngMgAZvnAOf3ACM/Abyb1PVqCPUMgPbt24d33nkHAPD1118jNDQUf/zxB7755hvMmDGDAVB1TAGQyjxPCpGzMuYB6Uuqbs5nCxA1JpYznO9dAhz6FgiMFIv/Xj0DZB+v+nV+4UD7QeL1Oekit+fK3+I1+hJ5WY0W6Hg30CtJ/AOhUsmPSxKw9V0g9RXRFfdXKvDf/sADy4DmPa15tS6hXgFQUVER/Pz8AAA///wz7rvvPqjVatx88804e/asVSvoVIyJc+7e1/9gEzmbG+UBsQWIGhPPACCkM5BZMbt5SS6QkVv18hjhMUCHu4AOg0XXWXV/7yVJBDLGh0oDaGr4WFapgH7TgIgewNePAkXZQG46sHQoMGkr0KRNQ6/SpdRrFFi7du2wZs0anDt3Dhs2bMCgQYMAAJmZmfD397dqBZ1K+wSg03CgwxCla0JkH60sh8NXmg+ILUDU2Iz8BOjxsJjnKigKULtXHFABLfsCiSnAtIPAPzYDA54HwrvX/M+uSiVyeDTuYgLRmoIfS236A//8HWhRMeq6rAjYu6whV+aS6tUCNGPGDDz44IN46qmncMcddyA+XkzT/vPPP6NHjx5WraBTGfS60jUgsq+ofsDmiu2zlfKA2AJEjU1oJ2D4B+bnBgNQmCmCF3uPZPSPAMZ+Abx1EyDpgUPfAANnVj26jKpUrzs1atQopKenY8+ePdiwYYNp/8CBA025QUREaNHbYj6gShMisgWIGju1GvALU24aB59goO0dYjv3HHBupzL1aKTqHSqGhYWhR48euHjxomll+D59+iA6OtpqlSOiRs7d07wu2NUzQM458zG2ABE1XNf7zdsHVylXj0aoXgGQwWDAq6++ioCAALRq1QqtWrVCYGAgXnvtNRgMBmvXkYgaM8vh8Gct8oDYAkTUcNF3AW4V06ocXi2WmKFaqVcO0Isvvoj//e9/mD17Nm65RSQ5btmyBbNmzUJxcTHeeOMNq1aSiBoxyzygM78D3ceIbWMLkEbLebGI6kvrJ4KgQ98A164Af/0K3GSHCYmNs1YfXg3oy0VOkn+4mBbALwzwCQF8mgHeTWuf3G1n9arVsmXL8Mknn5hWgQeAbt26oXnz5pg8eTIDICIyaxErFpLUlwJ//wYcXQvkXwLyRNc5W3+IGqjr/SIAAkQ3mC0DoOyTwN6lwJ9f1DxrtSWvIBEQ+RofoeKrX7j5HyIF1CsAunLlSpW5PtHR0bhy5UqDK0VETsTdS+QBnd0q5ixZOU5+nOuAETVM24Eij644Bzj2o1h30lrLLZUVi+TqvzcBf2268SzXVbl2VTwqTxbpHdz4AqDu3bvjgw8+wHvvvSfb/8EHH6Bbt27VvIrW/HEBmfnFUEGFx2/jhFXkQtrfKc//sdRlpH3rQuRs3DyAziNEy0xZEXB8HdB1VO1fX14CnPwZyDgMFOcBJXlASb5o4Tm/Byi/dv1rNB5Ax3uAXhPEsh15l4D8i+avhdniUZQNFGQChVny5aAA0QqkoHoFQHPnzsXQoUPxyy+/mOYA2r59O86dO4effvrJqhV0Jh9sOoVTmQXw1boxACLXEjcJKC8Vc6b4hVc8wsRkck3bKl07osav6/0iAAJEN9iNAiBJAs7vFl1Zh76VD0qoSVhXoPtYoNsYwKepeb9/BIAaFogFgJICoCBDBEQFGYBa2dyger17//79ceLECSxYsADHjh0DANx3332YOHEiXn/9ddx6661WraSz8HIXq/ZeK9NDkiSouBwGuQp3TzEzLhHZRsu+IgE57wJw6heg8LKYafr4OtEtlmsxBYVKBRRdFV3SN+IXAbS9HWhzu5iBuiGtNlpf8XCQf3pUkiRJ1jrZn3/+iZ49e0Kv11vrlIrIy8tDQEAAcnNzrbq0xwP/3Y5dp0WO1PHXB0PrprHauYmIyMX9/DKwrSI1JbgDcOUvwFB+49e5e4vurI53i5FbWj/zwzOwUa1dWZfPb8ccm+akvD3MAc+1Uj0DICIisp6u95sDoMoJx2o3QKUWXV+AWIMssg/Q/UGg4zDRMuNiGADZkWUAVFSqR6C3gpUhIiLnEtZVjLg8v1s8D2gJdLpHLMLdPJbrhFXCAMiOPN0tWoDKGnc3IRERORiVChizQkyGGHwTENGjUXVf2VudAqD77ruvxuM5OTkNqYvTq9wFRkREZFW+IYrOrdOY1CkACggIuOHx8ePHN6hCzszbw3y7ixgAERERKaZOAdCSJUtsVQ+X4OVumQNUi8x8IiIisglmRNmRl0UXWDFzgIiIiBTDAMiOKo8CIyIiImUoHgAtWLAAUVFR8PT0RFxcHHbt2lVt2cOHD2PkyJGIioqCSqXC/Pnzrysza9YsqFQq2aOqhVuVIO8CYwBERESkFEUDoJUrVyI5ORkzZ87Evn370L17dyQmJiIzM7PK8kVFRWjTpg1mz56NsLCwas/buXNnXLp0yfTYsmWLrS6hTtgFRkRE5BgUDYDmzZuHxx9/HElJSejUqRMWLVoEb29vLF68uMryvXv3xptvvokxY8ZAq9VWe143NzeEhYWZHsHBwba6hDphFxgREZFjUCwAKi0txd69e5GQkGCujFqNhIQEbN++vUHnPnnyJCIiItCmTRuMGzcO6ek1L/hWUlKCvLw82cMWvNw5DJ6IiMgRKBYAZWdnQ6/XIzQ0VLY/NDQUOp2u3ueNi4vD0qVLsX79eixcuBCnT5/Grbfeivz8/Gpfk5KSgoCAANMjMjKy3u9fE/lEiBwGT0REpBTFk6CtbciQIbj//vvRrVs3JCYm4qeffkJOTg6++uqral8zffp05Obmmh7nzp2zSd0sc4C4FAYREZFyFFsLLDg4GBqNBhkZGbL9GRkZNSY411VgYCBuuukmnDp1qtoyWq22xpwia+EoMCIiIsegWAuQh4cHevXqhdTUVNM+g8GA1NRUxMfHW+19CgoK8NdffyE8PNxq56wvrgVGRETkGBRdDT45ORkTJkxAbGws+vTpg/nz56OwsBBJSUkAgPHjx6N58+ZISUkBIBKnjxw5Ytq+cOEC9u/fD19fX7Rr1w4A8Mwzz2DYsGFo1aoVLl68iJkzZ0Kj0WDs2LHKXKQFrgVGRETkGBQNgEaPHo2srCzMmDEDOp0OMTExWL9+vSkxOj09HWq1uZHq4sWL6NGjh+n5W2+9hbfeegv9+/dHWloaAOD8+fMYO3YsLl++jGbNmqFfv37YsWMHmjVrZtdrq4rWzXwtzAEiIiJSjkqSJEnpSjiavLw8BAQEIDc3F/7+/lY9d8eX1+NamR4dQv2w4anbrHpuIiIiV1aXz2+nGwXm6Ix5QEVlHAZPRESkFAZAdmYcCs8kaCIiIuUwALIz41B4BkBERETKYQBkZ+YuMD2YfkVERKQMBkB2ZuwCkySgpNygcG2IiIhcEwMgO7OcDZrdYERERMpgAGRnsskQORcQERGRIhgA2ZkXV4QnIiJSHAMgO7NcD4zLYRARESmDAZCdMQeIiIhIeQyA7MyyC4w5QERERMpgAGRn3h5sASIiIlIaAyA787IcBcYAiIiISBEMgOxMlgPELjAiIiJFMACyM28OgyciIlIcAyA78+IweCIiIsUxALIzdoEREREpjwGQnXEUGBERkfIYANkZZ4ImIiJSHgMgO7McBs8WICIiImUwALIz5gAREREpjwGQncm7wDgMnoiISAkMgOxM66aGSiW22QVGRESkDAZAdqZSqeBd0Q3GJGgiIiJlMABSgHEyROYAERERKYMBkAJMARBbgIiIiBTBAEgB3u5iKDy7wIiIiJTBAEgBll1gBoOkcG2IiIhcDwMgBVjOBVRSblCwJkRERK6JAZACOBcQERGRshgAKcCL64EREREpigGQAiy7wIo5FJ6IiMjuGAApgCvCExERKYsBkAIsV4RnAERERGR/DIAUYNkCdK2MSdBERET2xgBIAZY5QNdKOQyeiIjI3hgAKcCLw+CJiIgUpXgAtGDBAkRFRcHT0xNxcXHYtWtXtWUPHz6MkSNHIioqCiqVCvPnz2/wOZUg7wJjDhAREZG9KRoArVy5EsnJyZg5cyb27duH7t27IzExEZmZmVWWLyoqQps2bTB79myEhYVZ5ZxK4CgwIiIiZSkaAM2bNw+PP/44kpKS0KlTJyxatAje3t5YvHhxleV79+6NN998E2PGjIFWq7XKOZXgKcsBYgBERERkb4oFQKWlpdi7dy8SEhLMlVGrkZCQgO3btzvMOW3B22IYPLvAiIiI7M/txkVsIzs7G3q9HqGhobL9oaGhOHbsmF3PWVJSgpKSEtPzvLy8er1/bXEtMCIiImUpngTtCFJSUhAQEGB6REZG2vT9PDkMnoiISFGKBUDBwcHQaDTIyMiQ7c/IyKg2wdlW55w+fTpyc3NNj3PnztXr/WuLEyESEREpS7EAyMPDA7169UJqaqppn8FgQGpqKuLj4+16Tq1WC39/f9nDljgKjIiISFmK5QABQHJyMiZMmIDY2Fj06dMH8+fPR2FhIZKSkgAA48ePR/PmzZGSkgJAJDkfOXLEtH3hwgXs378fvr6+aNeuXa3O6Qi8GAAREREpStEAaPTo0cjKysKMGTOg0+kQExOD9evXm5KY09PToVabG6kuXryIHj16mJ6/9dZbeOutt9C/f3+kpaXV6pyOwEOjhloFGCSgmKPAiIiI7E4lSZKkdCUcTV5eHgICApCbm2uz7rAuMzegoKQc7UJ88Utyf5u8BxERkSupy+c3R4EpxNgNxokQiYiI7I8BkEKMidCcB4iIiMj+GAApxKtiLiDOBE1ERGR/DIAUYuwCKy4zwGBgGhYREZE9MQBSiHwyRLYCERER2RMDIIV4uZtnIOBcQERERPbFAEghlpMhci4gIiIi+2IApBBvd84GTUREpBQGQAqRL4fBofBERET2xABIIV5MgiYiIlIMAyCFWHaBcTZoIiIi+2IApBCuCE9ERKQcBkAK8fYwD4NnCxAREZF9MQBSiJeH+dYzB4iIiMi+GAAphBMhEhERKYcBkEJkS2FwGDwREZFdMQBSiDeToImIiBTDAEghnu6cB4iIiEgpDIAUIu8CYwBERERkTwyAFGI5DJ5dYERERPbFAEghXpaLobILjIiIyK4YACnEciboYrYAERER2RUDIIV4uKnhplYBAIrKOAyeiIjInhgAKcjYCsQcICIiIvtiAKQgYx4Qu8CIiIjsiwGQgoxD4ZkETUREZF8MgBTkVTEUnl1gRERE9sUASEHGFqDScgP0Bknh2hAREbkOBkAK8uJyGERERIpgAKQgL9mCqBwKT0REZC8MgBTE9cCIiIiUwQBIQd6yFiAGQERERPbCAEhBnswBIiIiUgQDIAWxC4yIiEgZDIAU5F0xDxDALjAiIiJ7YgCkIHaBERERKYMBkILkXWAcBk9ERGQvDIAUxFFgREREynCIAGjBggWIioqCp6cn4uLisGvXrhrLr1q1CtHR0fD09ETXrl3x008/yY4/8sgjUKlUssfgwYNteQn1YjkTNAMgIiIi+1E8AFq5ciWSk5Mxc+ZM7Nu3D927d0diYiIyMzOrLL9t2zaMHTsWjz32GP744w+MGDECI0aMwKFDh2TlBg8ejEuXLpkeX3zxhT0up04sk6ALStgFRkREZC+KB0Dz5s3D448/jqSkJHTq1AmLFi2Ct7c3Fi9eXGX5d999F4MHD8azzz6Ljh074rXXXkPPnj3xwQcfyMpptVqEhYWZHkFBQfa4nDoJ8deatjPyihWsCRERkWtRNAAqLS3F3r17kZCQYNqnVquRkJCA7du3V/ma7du3y8oDQGJi4nXl09LSEBISgg4dOmDSpEm4fPlytfUoKSlBXl6e7GEP4QGepu1LOQyAiIiI7EXRACg7Oxt6vR6hoaGy/aGhodDpdFW+RqfT3bD84MGD8emnnyI1NRVz5szB5s2bMWTIEOj1VefZpKSkICAgwPSIjIxs4JXVjp+nO/y0ohvsUu41u7wnERERAW43LtL4jBkzxrTdtWtXdOvWDW3btkVaWhoGDhx4Xfnp06cjOTnZ9DwvL89uQVB4oCfyMwpwKbcYkiRBpVLZ5X2JiIhcmaItQMHBwdBoNMjIyJDtz8jIQFhYWJWvCQsLq1N5AGjTpg2Cg4Nx6tSpKo9rtVr4+/vLHvYSHuAFACgpN+BKYand3peIiMiVKRoAeXh4oFevXkhNTTXtMxgMSE1NRXx8fJWviY+Pl5UHgI0bN1ZbHgDOnz+Py5cvIzw83DoVtyJZHlAu84CIiIjsQfFRYMnJyfj444+xbNkyHD16FJMmTUJhYSGSkpIAAOPHj8f06dNN5adOnYr169fj7bffxrFjxzBr1izs2bMHU6ZMAQAUFBTg2WefxY4dO3DmzBmkpqZi+PDhaNeuHRITExW5xpoYW4AA4GIO84CIiIjsQfEcoNGjRyMrKwszZsyATqdDTEwM1q9fb0p0Tk9Ph1ptjtP69u2LFStW4KWXXsK///1vtG/fHmvWrEGXLl0AABqNBgcOHMCyZcuQk5ODiIgIDBo0CK+99hq0Wm2VdVBSeKC5BUjHofBERER2oZIkSVK6Eo4mLy8PAQEByM3NtXk+0JaT2XjofzsBAP/s3xYvDIm26fsRERE5q7p8fiveBebqLFuAOBSeiIjIPhgAKSzCIgeIkyESERHZBwMghXl5aBDo7Q4AuMgWICIiIrtgAOQAwvxFN1hGXjEMBqZkERER2RoDIAcQESi6wcr0ErILShSuDRERkfNjAOQALCdDvMjJEImIiGyOAZADMLYAAYCOeUBEREQ2xwDIAchagDgSjIiIyOYYADkAy+UwOBcQERGR7TEAcgARgcwBIiIisicGQA4g1N9iNmguiEpERGRzDIAcgKe7Bk19PAAAl9gCREREZHMMgByEcU2wjLxilOsNCteGiIjIuTEAchDGRGiDBGTmczJEIiIiW2IA5CAiAixXhWc3GBERkS0xAHIQ4YEcCk9ERGQvDIAchOVkiJc4GSIREZFNMQByEJaTIV5kCxAREZFNMQByEGwBIiIish8GQA4iLMATKpXYZg4QERGRbTEAchDuGjWa+WoBcDkMIiIiW2MA5ECMI8GyC0pQWs7JEImIiGyFAZADMc4FJEliRmgiIiKyDQZADiSMkyESERHZBQMgBxIRwMkQiYiI7IEBkAMxLogKABc5FJ6IiMhmGAA5kHC2ABEREdkFAyAHEsEWICIiIrtgAORAQvw8oVGL2RDZAkRERGQ7DIAciEatQqifmAyRo8CIiIhshwGQgzEOhb9SWIq/swoUrg0REZFzYgDkYHq3bmLanvT5PhSVlitYGyIiIufEAMjB/OuO9mgX4gsAOJ6Rj+nfHoQkSQrXioiIyLkwAHIwPlo3LHqoF3w8NACA7/ZfxLJtZ5StFBERkZNhAOSA2oX44q37u5uev/7jUew9e0XBGhERETkXBkAOakjXcPzjtjYAgHKDhEmf78Pes1fZHUZERGQFDIAc2LOJHXBzG5EUnZlfgpELt2HgvM1YsOkULuZwniAiIqL6cogAaMGCBYiKioKnpyfi4uKwa9euGsuvWrUK0dHR8PT0RNeuXfHTTz/JjkuShBkzZiA8PBxeXl5ISEjAyZMnbXkJNuGmUeP9sT3Rqqm3ad/fWYV4c8Nx3DLnVwx973c8/dWf+Pi3v/HbiSzocouhN7CFiIiI6EZUksJ9KitXrsT48eOxaNEixMXFYf78+Vi1ahWOHz+OkJCQ68pv27YNt912G1JSUnD33XdjxYoVmDNnDvbt24cuXboAAObMmYOUlBQsW7YMrVu3xssvv4yDBw/iyJEj8PT0vO6cleXl5SEgIAC5ubnw9/e3+jXXVWFJOX48eAnf7D2PnadrzgVSq4CmvloE+2rRzE+LIG93+Grd4OvpBj+tG3y0bvD20MDT3fzw0KjhplFBo1bBTS2+umvUcNeo4aZWwcNNfHVTX19OpVLZ6S4QERHVrC6f34oHQHFxcejduzc++OADAIDBYEBkZCSefPJJvPDCC9eVHz16NAoLC7F27VrTvptvvhkxMTFYtGgRJElCREQEnn76aTzzzDMAgNzcXISGhmLp0qUYM2bMDevkaAGQpXNXirD6jwtYe+AiTmUWQOkGH3dNRWCkVkGjUcEYDhkDI7VKbGtUKqhVgLoicNKoVGJbpULlGEplLFtxzPhcnEPsU6tUkCCJ65cAQ8WPsVptfq1xWZGqfsKN76lSiTqbnlfaV/l4VcxXLUgwv6G64jrVKhU0FfVCFedSQSWrg0ESLZkGSTJ9j433UpzC+BWy51XWz7RfXkZV6XjlOhi/hyqV+R4aJAmSJL/Gmq5Fb5BM16I3SFCpAI3x50UtyukNEsoNEsr1BpTrJUCFiuNquFcE3MZ7bFl/Yz0k0/1RQaOG7OdKbxB1LtdLpp8RjfFnsOJnRZzHfD6YnhvrLn8fSZIqfibl7weYz1P5/lt+nyozSJLs+21+nUVpi++RsQ56gwR9xffDIEmm75m64vfF+HsC4+9SRf2M30ODwfw7Y6yb8XU1fR8N1dwXI+PvqOX3yfJemH+3r/+Zrep7YHk/LG7FdT+Blf+uSKi4Rxb3yfT7XVGP6773FfW3/HtlrpeoT3WfmObvs8r0O2MqKkmynw3jexnvufFvROWfDsu6SRX3/7p9VVfHdC+Mv2uA+J6XG4zfQ8n0/jD+zFj8XTFek0ES339JMt/Hyn+bLe+R7J7A/HNvWV/jfQzwcsfjFbmu1lKXz283q75zHZWWlmLv3r2YPn26aZ9arUZCQgK2b99e5Wu2b9+O5ORk2b7ExESsWbMGAHD69GnodDokJCSYjgcEBCAuLg7bt2+vVQDkyCKbeONfA9vjXwPbo7hMj5MZBTimy8NxXT5OZxciq6AEWfniUW6H6KhML6FMr7f5+xARkXOJbOJl9QCoLhQNgLKzs6HX6xEaGirbHxoaimPHjlX5Gp1OV2V5nU5nOm7cV12ZykpKSlBSUmJ6npeXV7cLUYinuwZdWwSga4uA644ZDBJyrpUh71oZCkrKkV9cjsKSchSUlKO4TI9rFY/iUj1K9RL0BgP0BkBvMKCs4r9xEdyI/8rL9AaUV/w3Zdy2/K+93GAw5R8Zwy7jfymGiv/C9RX/eRj/KzP9Z1Hp/wbLVh3LFhAiInIeBoOy769oAOQoUlJS8MorryhdDatSq1Vo4uOBJj4eSlfFKozdEaagyGDsooKs6diyjOhykTebAxZNsKhokq7UnFy5mbmqNubKuyTp+m4y4+v1BnMQaNnFIa+P+ZkkVXRLQN6dYKyfsQsDsjpXc99MTedVd1tZvs6yeb7yvbCsS03dbZXPY+yKNHbLSIApADY2xVvml7lVtKeLQLsi4K74Kym/RlEry+Z64/de3G9Rb7VKBTeNylQPyfSzAVNAroK4QHm3omV3kHy/sSvJMqA3GCRZN5X8e2bu0qnqXqnVFV2lKvP7yLpLpOv/sVCrRJeP2vLeWvx8GK/f2D1m7D4xdS+rzb8zEsQ/TMYuLlR+f0gVXYrmbkt1xX0xdctC/vMizlPVD0nF77Gh+n9sKnclGc9tup+Sxe+1xX227O4ymLpbRXeWMV/RWD/jPan8vYfFMeO5KorU+PNfuUvKskvSsjupcjez5c+RqSuz0rlVkHdjVu4CN9ZffpfF+corfjbLK7qwLH8X1BVvZNkFa/l3wvh3w/jzqbHoprO8h3rj71Cl663882Cqs0V3m6e7suOwFA2AgoODodFokJGRIdufkZGBsLCwKl8TFhZWY3nj14yMDISHh8vKxMTEVHnO6dOny7rV8vLyEBkZWefrIdtRVeTQaKr8o0pERFQ3ioZfHh4e6NWrF1JTU037DAYDUlNTER8fX+Vr4uPjZeUBYOPGjabyrVu3RlhYmKxMXl4edu7cWe05tVot/P39ZQ8iIiJyXop3gSUnJ2PChAmIjY1Fnz59MH/+fBQWFiIpKQkAMH78eDRv3hwpKSkAgKlTp6J///54++23MXToUHz55ZfYs2cPPvroIwCipWDatGl4/fXX0b59e9Mw+IiICIwYMUKpyyQiIiIHongANHr0aGRlZWHGjBnQ6XSIiYnB+vXrTUnM6enpUKvNDVV9+/bFihUr8NJLL+Hf//432rdvjzVr1pjmAAKA5557DoWFhZg4cSJycnLQr18/rF+/vlZzABEREZHzU3weIEfkyPMAERERUdXq8vntEEthEBEREdkTAyAiIiJyOQyAiIiIyOUwACIiIiKXwwCIiIiIXA4DICIiInI5DICIiIjI5TAAIiIiIpfDAIiIiIhcjuJLYTgi4+TYeXl5CteEiIiIasv4uV2bRS4YAFUhPz8fABAZGalwTYiIiKiu8vPzERAQUGMZrgVWBYPBgIsXL8LPzw8qlcqq587Ly0NkZCTOnTvHdcZsjPfafniv7Yf32n54r+3HWvdakiTk5+cjIiJCtpB6VdgCVAW1Wo0WLVrY9D38/f35C2UnvNf2w3ttP7zX9sN7bT/WuNc3avkxYhI0ERERuRwGQERERORyGADZmVarxcyZM6HVapWuitPjvbYf3mv74b22H95r+1HiXjMJmoiIiFwOW4CIiIjI5TAAIiIiIpfDAIiIiIhcDgMgIiIicjkMgOxowYIFiIqKgqenJ+Li4rBr1y6lq9TopaSkoHfv3vDz80NISAhGjBiB48ePy8oUFxfjiSeeQNOmTeHr64uRI0ciIyNDoRo7j9mzZ0OlUmHatGmmfbzX1nPhwgU89NBDaNq0Kby8vNC1a1fs2bPHdFySJMyYMQPh4eHw8vJCQkICTp48qWCNGye9Xo+XX34ZrVu3hpeXF9q2bYvXXntNtpYU73X9/Pbbbxg2bBgiIiKgUqmwZs0a2fHa3NcrV65g3Lhx8Pf3R2BgIB577DEUFBRYpX4MgOxk5cqVSE5OxsyZM7Fv3z50794diYmJyMzMVLpqjdrmzZvxxBNPYMeOHdi4cSPKysowaNAgFBYWmso89dRT+OGHH7Bq1Sps3rwZFy9exH333adgrRu/3bt347///S+6desm2897bR1Xr17FLbfcAnd3d6xbtw5HjhzB22+/jaCgIFOZuXPn4r333sOiRYuwc+dO+Pj4IDExEcXFxQrWvPGZM2cOFi5ciA8++ABHjx7FnDlzMHfuXLz//vumMrzX9VNYWIju3btjwYIFVR6vzX0dN24cDh8+jI0bN2Lt2rX47bffMHHiROtUUCK76NOnj/TEE0+Ynuv1eikiIkJKSUlRsFbOJzMzUwIgbd68WZIkScrJyZHc3d2lVatWmcocPXpUAiBt375dqWo2avn5+VL79u2ljRs3Sv3795emTp0qSRLvtTU9//zzUr9+/ao9bjAYpLCwMOnNN9807cvJyZG0Wq30xRdf2KOKTmPo0KHSo48+Ktt33333SePGjZMkiffaWgBIq1evNj2vzX09cuSIBEDavXu3qcy6desklUolXbhwocF1YguQHZSWlmLv3r1ISEgw7VOr1UhISMD27dsVrJnzyc3NBQA0adIEALB3716UlZXJ7n10dDRatmzJe19PTzzxBIYOHSq7pwDvtTV9//33iI2Nxf3334+QkBD06NEDH3/8sen46dOnodPpZPc6ICAAcXFxvNd11LdvX6SmpuLEiRMAgD///BNbtmzBkCFDAPBe20pt7uv27dsRGBiI2NhYU5mEhASo1Wrs3LmzwXXgYqh2kJ2dDb1ej9DQUNn+0NBQHDt2TKFaOR+DwYBp06bhlltuQZcuXQAAOp0OHh4eCAwMlJUNDQ2FTqdToJaN25dffol9+/Zh9+7d1x3jvbaev//+GwsXLkRycjL+/e9/Y/fu3fjXv/4FDw8PTJgwwXQ/q/qbwntdNy+88ALy8vIQHR0NjUYDvV6PN954A+PGjQMA3msbqc191el0CAkJkR13c3NDkyZNrHLvGQCR03jiiSdw6NAhbNmyRemqOKVz585h6tSp2LhxIzw9PZWujlMzGAyIjY3Ff/7zHwBAjx49cOjQISxatAgTJkxQuHbO5auvvsLy5cuxYsUKdO7cGfv378e0adMQERHBe+3k2AVmB8HBwdBoNNeNhsnIyEBYWJhCtXIuU6ZMwdq1a7Fp0ya0aNHCtD8sLAylpaXIycmRlee9r7u9e/ciMzMTPXv2hJubG9zc3LB582a89957cHNzQ2hoKO+1lYSHh6NTp06yfR07dkR6ejoAmO4n/6Y03LPPPosXXngBY8aMQdeuXfHwww/jqaeeQkpKCgDea1upzX0NCwu7bqBQeXk5rly5YpV7zwDIDjw8PNCrVy+kpqaa9hkMBqSmpiI+Pl7BmjV+kiRhypQpWL16NX799Ve0bt1adrxXr15wd3eX3fvjx48jPT2d976OBg4ciIMHD2L//v2mR2xsLMaNG2fa5r22jltuueW66RxOnDiBVq1aAQBat26NsLAw2b3Oy8vDzp07ea/rqKioCGq1/KNQo9HAYDAA4L22ldrc1/j4eOTk5GDv3r2mMr/++isMBgPi4uIaXokGp1FTrXz55ZeSVquVli5dKh05ckSaOHGiFBgYKOl0OqWr1qhNmjRJCggIkNLS0qRLly6ZHkVFRaYy//znP6WWLVtKv/76q7Rnzx4pPj5eio+PV7DWzsNyFJgk8V5by65duyQ3NzfpjTfekE6ePCktX75c8vb2lj7//HNTmdmzZ0uBgYHSd999Jx04cEAaPny41Lp1a+natWsK1rzxmTBhgtS8eXNp7dq10unTp6Vvv/1WCg4Olp577jlTGd7r+snPz5f++OMP6Y8//pAASPPmzZP++OMP6ezZs5Ik1e6+Dh48WOrRo4e0c+dOacuWLVL79u2lsWPHWqV+DIDs6P3335datmwpeXh4SH369JF27NihdJUaPQBVPpYsWWIqc+3aNWny5MlSUFCQ5O3tLd17773SpUuXlKu0E6kcAPFeW88PP/wgdenSRdJqtVJ0dLT00UcfyY4bDAbp5ZdflkJDQyWtVisNHDhQOn78uEK1bbzy8vKkqVOnSi1btpQ8PT2lNm3aSC+++KJUUlJiKsN7XT+bNm2q8u/zhAkTJEmq3X29fPmyNHbsWMnX11fy9/eXkpKSpPz8fKvUTyVJFtNdEhEREbkA5gARERGRy2EARERERC6HARARERG5HAZARERE5HIYABEREZHLYQBERERELocBEBEREbkcBkBERLWgUqmwZs0apatBRFbCAIiIHN4jjzwClUp13WPw4MFKV42IGik3pStARFQbgwcPxpIlS2T7tFqtQrUhosaOLUBE1ChotVqEhYXJHkFBQQBE99TChQsxZMgQeHl5oU2bNvj6669lrz948CDuuOMOeHl5oWnTppg4cSIKCgpkZRYvXozOnTtDq9UiPDwcU6ZMkR3Pzs7GvffeC29vb7Rv3x7ff/+9bS+aiGyGARAROYWXX34ZI0eOxJ9//olx48ZhzJgxOHr0KACgsLAQiYmJCAoKwu7du7Fq1Sr88ssvsgBn4cKFeOKJJzBx4kQcPHgQ33//Pdq1ayd7j1deeQUPPPAADhw4gLvuugvjxo3DlStX7HqdRGQlVllSlYjIhiZMmCBpNBrJx8dH9njjjTckSZIkANI///lP2Wvi4uKkSZMmSZIkSR999JEUFBQkFRQUmI7/+OOPklqtlnQ6nSRJkhQRESG9+OKL1dYBgPTSSy+ZnhcUFEgApHXr1lntOonIfpgDRESNwu23346FCxfK9jVp0sS0HR8fLzsWHx+P/fv3AwCOHj2K7t27w8fHx3T8lltugcFgwPHjx6FSqXDx4kUMHDiwxjp069bNtO3j4wN/f39kZmbW95KISEEMgIioUfDx8bmuS8pavLy8alXO3d1d9lylUsFgMNiiSkRkY8wBIiKnsGPHjuued+zYEQDQsWNH/PnnnygsLDQd37p1K9RqNTp06AA/Pz9ERUUhNTXVrnUmIuWwBYiIGoWSkhLodDrZPjc3NwQHBwMAVq1ahdjYWPTr1w/Lly/Hrl278L///Q8AMG7cOMycORMTJkzArFmzkJWVhSeffBIPP/wwQkNDAQCzZs3CP//5T4SEhGDIkCHIz8/H1q1b8eSTT9r3QonILhgAEVGjsH79eoSHh8v2dejQAceOHQMgRmh9+eWXmDx5MsLDw/HFF1+gU6dOAABvb29s2LABU6dORe/eveHt7Y2RI0di3rx5pnNNmDABxcXFeOedd/DMM88gODgYo0aNst8FEpFdqSRJkpSuBBFRQ6hUKqxevRojRoxQuipE1EgwB4iIiIhcDgMgIiIicjnMASKiRo89+URUV2wBIiIiIpfDAIiIiIhcDgMgIiIicjkMgIiIiMjlMAAiIiIil8MAiIiIiFwOAyAiIiJyOQyAiIiIyOUwACIiIiKX8/9tKwc4jb3SzwAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rRXjw--5YxK-",
        "outputId": "ee93a095-74ac-4fec-f77b-14db2cbc845f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_4\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_6 (InputLayer)        [(None, 50)]              0         \n",
            "                                                                 \n",
            " dense_13 (Dense)            (None, 64)                3264      \n",
            "                                                                 \n",
            " dropout_8 (Dropout)         (None, 64)                0         \n",
            "                                                                 \n",
            " batch_normalization_8 (Batc  (None, 64)               256       \n",
            " hNormalization)                                                 \n",
            "                                                                 \n",
            " dense_14 (Dense)            (None, 50)                3250      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 6,770\n",
            "Trainable params: 6,642\n",
            "Non-trainable params: 128\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Anomaly detection\n",
        "anomaly_scores = []\n",
        "for inputs in val_data2_scaled:\n",
        "    inputs = tf.expand_dims(inputs, 0)  # Add batch dimension\n",
        "    outputs = model.predict(inputs)\n",
        "    loss = loss_fn(outputs, inputs)\n",
        "    anomaly_scores.append(loss.numpy())\n",
        "\n",
        "# Thresholding and labeling anomalies\n",
        "#threshold = 0.10  # Adjust threshold based on your data and requirements\n",
        "#anomalies = [1 if score > threshold else 0 for score in anomaly_scores]\n",
        "\n",
        "# Use anomalies as needed (e.g., visualization, further analysis)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FYJe1QHs5ZKC",
        "outputId": "f7aa5eb1-3137-42e2-d511-67dcc7f63e3c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 71ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 50ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 118ms/step\n",
            "1/1 [==============================] - 0s 79ms/step\n",
            "1/1 [==============================] - 0s 119ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 60ms/step\n",
            "1/1 [==============================] - 0s 191ms/step\n",
            "1/1 [==============================] - 0s 238ms/step\n",
            "1/1 [==============================] - 0s 56ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.hist(anomaly_scores)\n",
        "plt.show"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 447
        },
        "id": "rAej042DGwqV",
        "outputId": "ae12affb-539e-4de0-b54b-8ae7a3395f7b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<function matplotlib.pyplot.show(close=None, block=None)>"
            ]
          },
          "metadata": {},
          "execution_count": 50
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.percentile(anomaly_scores, 90)\n",
        "threshold"
      ],
      "metadata": {
        "id": "2CoO5NXV84SO",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3a281171-0f62-47d1-fd52-703cc19c1ea6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.313055466691831"
            ]
          },
          "metadata": {},
          "execution_count": 51
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.mean(anomaly_scores) + np.std(anomaly_scores)\n",
        "threshold"
      ],
      "metadata": {
        "id": "jzuN2fszAcIl",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9dcb558c-fb4c-469d-f302-d2bbe5758f83"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.25068553278870576"
            ]
          },
          "metadata": {},
          "execution_count": 165
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.percentile(anomaly_scores, 75) + (np.std(anomaly_scores)/np.sqrt(len(anomaly_scores)))"
      ],
      "metadata": {
        "id": "YMsUDdWuHaf2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "threshold"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "npPhkI76qbkr",
        "outputId": "bb0b8cfd-e4b3-46c4-f791-35861b68b5ac"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.33162285956700444"
            ]
          },
          "metadata": {},
          "execution_count": 109
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "anomaly_scores = []\n",
        "for inputs in test_data_scaled:\n",
        "    inputs = tf.expand_dims(inputs, 0)  # Add batch dimension\n",
        "    outputs = model.predict(inputs)\n",
        "    loss = loss_fn(outputs, inputs)\n",
        "    anomaly_scores.append(loss.numpy())"
      ],
      "metadata": {
        "id": "BdIgv1zCxsO9",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4fe0e140-da6a-4f86-c8d7-6532793250c3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 48ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "anomalies = [1 if score > threshold else 0 for score in anomaly_scores]\n",
        "anomalies"
      ],
      "metadata": {
        "id": "l9swSGIkyCzf",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6d6fbd69-d1d4-4b16-d26b-c54dd972391b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 0]"
            ]
          },
          "metadata": {},
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report #hiddendim=64 1layer\n",
        "print(classification_report(gt, anomalies, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9Fk4SBpo2j0r",
        "outputId": "d4f76f51-46cf-4047-9260-387fffc18541"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.58      0.97      0.73        71\n",
            "      attack       0.00      0.00      0.00        49\n",
            "\n",
            "    accuracy                           0.57       120\n",
            "   macro avg       0.29      0.49      0.37       120\n",
            "weighted avg       0.35      0.57      0.43       120\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report #hiddendim=128 1layer\n",
        "print(classification_report(gt, anomalies, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "q1B9N2VM2HCq",
        "outputId": "091d616a-fd0e-4fc7-cc60-ff3d1d560cc3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.66      0.94      0.78        71\n",
            "      attack       0.79      0.31      0.44        49\n",
            "\n",
            "    accuracy                           0.68       120\n",
            "   macro avg       0.73      0.62      0.61       120\n",
            "weighted avg       0.71      0.68      0.64       120\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report #hiddendim=128 2layers\n",
        "print(classification_report(gt, anomalies, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YIstt6_B1tWU",
        "outputId": "bc78c9b1-41e8-49b4-9719-315ee7b779f0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.71      0.62      0.66        71\n",
            "      attack       0.53      0.63      0.58        49\n",
            "\n",
            "    accuracy                           0.62       120\n",
            "   macro avg       0.62      0.63      0.62       120\n",
            "weighted avg       0.64      0.62      0.63       120\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report #hiddendim=64 2layers\n",
        "print(classification_report(gt, anomalies, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fJAlLo_W0p_B",
        "outputId": "719c3416-942b-48fe-aacd-900c48e02d7e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.66      0.96      0.78        71\n",
            "      attack       0.82      0.29      0.42        49\n",
            "\n",
            "    accuracy                           0.68       120\n",
            "   macro avg       0.74      0.62      0.60       120\n",
            "weighted avg       0.73      0.68      0.64       120\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras.models import Model\n",
        "from tensorflow.keras.layers import Input, Dense, Dropout, BatchNormalization, concatenate\n",
        "\n",
        "def UNet(input_dim, hidden_dim, output_dim):\n",
        "    inputs = Input(shape=(input_dim,))\n",
        "\n",
        "    # Encoder\n",
        "    encoder1 = Dense(hidden_dim, activation='relu')(inputs)\n",
        "    encoder1 = Dropout(0.2)(encoder1)\n",
        "    encoder1 = BatchNormalization()(encoder1)\n",
        "\n",
        "    encoder2 = Dense(hidden_dim/2, activation='relu')(encoder1)\n",
        "    encoder2 = Dropout(0.2)(encoder2)\n",
        "    encoder2 = BatchNormalization()(encoder2)\n",
        "\n",
        "    # Decoder\n",
        "    decoder2 = Dense(hidden_dim/2, activation='relu')(encoder2)\n",
        "    decoder2 = Dropout(0.2)(decoder2)\n",
        "    decoder2 = BatchNormalization()(decoder2)\n",
        "    decoder2 = concatenate([encoder2, decoder2])  # Skip connection\n",
        "\n",
        "    decoder1 = Dense(hidden_dim, activation='relu')(decoder2)\n",
        "    decoder1 = Dropout(0.2)(decoder1)\n",
        "    decoder1 = BatchNormalization()(decoder1)\n",
        "    decoder1 = concatenate([encoder1, decoder1])  # Skip connection\n",
        "\n",
        "    # Output\n",
        "    output = Dense(output_dim, activation='sigmoid')(decoder1)\n",
        "\n",
        "    model = Model(inputs=inputs, outputs=output)\n",
        "    return model\n",
        "\n",
        "# Define your dataset and data preprocessing\n",
        "# ...\n",
        "\n",
        "# Define hyperparameters\n",
        "input_dim = train_data_scaled.shape[1]\n",
        "hidden_dim = 128\n",
        "output_dim = train_data_scaled.shape[1]\n",
        "learning_rate = 0.001\n",
        "batch_size = 64\n",
        "num_epochs = 100\n",
        "\n",
        "# Instantiate the UNet model\n",
        "model = UNet(input_dim, hidden_dim, output_dim)\n",
        "\n",
        "# Define loss function and optimizer\n",
        "loss_fn = tf.keras.losses.MeanSquaredError()\n",
        "optimizer = tf.keras.optimizers.Adam(learning_rate=learning_rate)\n",
        "\n",
        "# Compile the model\n",
        "model.compile(optimizer=optimizer, loss=loss_fn)\n",
        "\n",
        "# Train the model\n",
        "model.fit(train_data_scaled, train_data_scaled, epochs=num_epochs, batch_size=batch_size, validation_data=(val_data1_scaled, val_data1_scaled))\n"
      ],
      "metadata": {
        "id": "lThg9wLPyZ2o",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d2a3da95-abac-4644-c784-02badc9dc19c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "71/71 [==============================] - 6s 9ms/step - loss: 0.1585 - val_loss: 0.2147\n",
            "Epoch 2/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 0.0189 - val_loss: 0.3100\n",
            "Epoch 3/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0043 - val_loss: 0.3184\n",
            "Epoch 4/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0028 - val_loss: 0.3151\n",
            "Epoch 5/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0020 - val_loss: 0.3087\n",
            "Epoch 6/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0017 - val_loss: 0.2963\n",
            "Epoch 7/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 0.0015 - val_loss: 0.2417\n",
            "Epoch 8/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0013 - val_loss: 0.2296\n",
            "Epoch 9/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0012 - val_loss: 0.2154\n",
            "Epoch 10/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0011 - val_loss: 0.1937\n",
            "Epoch 11/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0010 - val_loss: 0.1938\n",
            "Epoch 12/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0010 - val_loss: 0.2204\n",
            "Epoch 13/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 9.3858e-04 - val_loss: 0.1955\n",
            "Epoch 14/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 8.7938e-04 - val_loss: 0.2095\n",
            "Epoch 15/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 8.5951e-04 - val_loss: 0.1758\n",
            "Epoch 16/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 8.6270e-04 - val_loss: 0.1783\n",
            "Epoch 17/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 7.6193e-04 - val_loss: 0.1759\n",
            "Epoch 18/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 7.6737e-04 - val_loss: 0.2279\n",
            "Epoch 19/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 7.2399e-04 - val_loss: 0.1522\n",
            "Epoch 20/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 7.2553e-04 - val_loss: 0.1454\n",
            "Epoch 21/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 6.6461e-04 - val_loss: 0.1756\n",
            "Epoch 22/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 6.9380e-04 - val_loss: 0.1446\n",
            "Epoch 23/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 6.7238e-04 - val_loss: 0.1400\n",
            "Epoch 24/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 6.5500e-04 - val_loss: 0.1397\n",
            "Epoch 25/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 6.2802e-04 - val_loss: 0.1372\n",
            "Epoch 26/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 6.5317e-04 - val_loss: 0.1367\n",
            "Epoch 27/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 6.2331e-04 - val_loss: 0.1380\n",
            "Epoch 28/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 6.2090e-04 - val_loss: 0.1381\n",
            "Epoch 29/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 6.0406e-04 - val_loss: 0.1416\n",
            "Epoch 30/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 6.1361e-04 - val_loss: 0.1403\n",
            "Epoch 31/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 6.4850e-04 - val_loss: 0.1406\n",
            "Epoch 32/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 6.0281e-04 - val_loss: 0.1412\n",
            "Epoch 33/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 6.0269e-04 - val_loss: 0.1410\n",
            "Epoch 34/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.6172e-04 - val_loss: 0.1359\n",
            "Epoch 35/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 6.1226e-04 - val_loss: 0.1360\n",
            "Epoch 36/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.9797e-04 - val_loss: 0.1336\n",
            "Epoch 37/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.8847e-04 - val_loss: 0.2241\n",
            "Epoch 38/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.9240e-04 - val_loss: 0.1632\n",
            "Epoch 39/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.9381e-04 - val_loss: 0.1344\n",
            "Epoch 40/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.4786e-04 - val_loss: 0.1340\n",
            "Epoch 41/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.9878e-04 - val_loss: 0.1331\n",
            "Epoch 42/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.9963e-04 - val_loss: 0.1261\n",
            "Epoch 43/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.4934e-04 - val_loss: 0.1129\n",
            "Epoch 44/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.5642e-04 - val_loss: 0.1120\n",
            "Epoch 45/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.8558e-04 - val_loss: 0.2193\n",
            "Epoch 46/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.6719e-04 - val_loss: 0.1550\n",
            "Epoch 47/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.8243e-04 - val_loss: 0.1482\n",
            "Epoch 48/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 6.0110e-04 - val_loss: 0.1392\n",
            "Epoch 49/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 5.1869e-04 - val_loss: 0.1334\n",
            "Epoch 50/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 5.3706e-04 - val_loss: 0.1352\n",
            "Epoch 51/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 5.2628e-04 - val_loss: 0.1378\n",
            "Epoch 52/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 5.2589e-04 - val_loss: 0.1399\n",
            "Epoch 53/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 5.2554e-04 - val_loss: 0.1491\n",
            "Epoch 54/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 4.9836e-04 - val_loss: 0.1437\n",
            "Epoch 55/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 5.2978e-04 - val_loss: 0.1420\n",
            "Epoch 56/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.0726e-04 - val_loss: 0.1502\n",
            "Epoch 57/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.2857e-04 - val_loss: 0.1885\n",
            "Epoch 58/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.3908e-04 - val_loss: 0.1392\n",
            "Epoch 59/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.9202e-04 - val_loss: 0.1758\n",
            "Epoch 60/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 4.8990e-04 - val_loss: 0.1496\n",
            "Epoch 61/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.1364e-04 - val_loss: 0.1322\n",
            "Epoch 62/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.2985e-04 - val_loss: 0.1315\n",
            "Epoch 63/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.0838e-04 - val_loss: 0.1315\n",
            "Epoch 64/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.0685e-04 - val_loss: 0.1307\n",
            "Epoch 65/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.9383e-04 - val_loss: 0.1348\n",
            "Epoch 66/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 4.8707e-04 - val_loss: 0.1353\n",
            "Epoch 67/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 5.0206e-04 - val_loss: 0.1663\n",
            "Epoch 68/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.0574e-04 - val_loss: 0.1602\n",
            "Epoch 69/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.9809e-04 - val_loss: 0.1659\n",
            "Epoch 70/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.8922e-04 - val_loss: 0.1435\n",
            "Epoch 71/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.1806e-04 - val_loss: 0.1467\n",
            "Epoch 72/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.9794e-04 - val_loss: 0.1326\n",
            "Epoch 73/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 4.5407e-04 - val_loss: 0.1361\n",
            "Epoch 74/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 4.8464e-04 - val_loss: 0.1341\n",
            "Epoch 75/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 4.7873e-04 - val_loss: 0.1350\n",
            "Epoch 76/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 4.7118e-04 - val_loss: 0.1319\n",
            "Epoch 77/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.7690e-04 - val_loss: 0.1310\n",
            "Epoch 78/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.8181e-04 - val_loss: 0.1323\n",
            "Epoch 79/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 5.0987e-04 - val_loss: 0.1301\n",
            "Epoch 80/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.3773e-04 - val_loss: 0.1329\n",
            "Epoch 81/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 4.4948e-04 - val_loss: 0.1387\n",
            "Epoch 82/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 4.6069e-04 - val_loss: 0.1433\n",
            "Epoch 83/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 4.3722e-04 - val_loss: 0.1375\n",
            "Epoch 84/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 4.6062e-04 - val_loss: 0.1225\n",
            "Epoch 85/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 5.3918e-04 - val_loss: 0.1244\n",
            "Epoch 86/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 4.8830e-04 - val_loss: 0.1241\n",
            "Epoch 87/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 4.2762e-04 - val_loss: 0.1291\n",
            "Epoch 88/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 4.4592e-04 - val_loss: 0.1230\n",
            "Epoch 89/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.6102e-04 - val_loss: 0.2051\n",
            "Epoch 90/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.3009e-04 - val_loss: 0.2142\n",
            "Epoch 91/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 4.3095e-04 - val_loss: 0.1886\n",
            "Epoch 92/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.3497e-04 - val_loss: 0.1942\n",
            "Epoch 93/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.1843e-04 - val_loss: 0.1708\n",
            "Epoch 94/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.3359e-04 - val_loss: 0.1716\n",
            "Epoch 95/100\n",
            "71/71 [==============================] - 0s 5ms/step - loss: 4.4076e-04 - val_loss: 0.1426\n",
            "Epoch 96/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.3048e-04 - val_loss: 0.1533\n",
            "Epoch 97/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.1562e-04 - val_loss: 0.1454\n",
            "Epoch 98/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.2841e-04 - val_loss: 0.1650\n",
            "Epoch 99/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.3258e-04 - val_loss: 0.1478\n",
            "Epoch 100/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 4.3418e-04 - val_loss: 0.1416\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7ff234dac220>"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Dbe6vmiXvyAg",
        "outputId": "6942b7a1-eb5a-465c-a13a-717dc5d00adf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_3\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                   Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            " input_4 (InputLayer)           [(None, 50)]         0           []                               \n",
            "                                                                                                  \n",
            " dense_15 (Dense)               (None, 128)          6528        ['input_4[0][0]']                \n",
            "                                                                                                  \n",
            " dropout_12 (Dropout)           (None, 128)          0           ['dense_15[0][0]']               \n",
            "                                                                                                  \n",
            " batch_normalization_12 (BatchN  (None, 128)         512         ['dropout_12[0][0]']             \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " dense_16 (Dense)               (None, 64)           8256        ['batch_normalization_12[0][0]'] \n",
            "                                                                                                  \n",
            " dropout_13 (Dropout)           (None, 64)           0           ['dense_16[0][0]']               \n",
            "                                                                                                  \n",
            " batch_normalization_13 (BatchN  (None, 64)          256         ['dropout_13[0][0]']             \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " dense_17 (Dense)               (None, 64)           4160        ['batch_normalization_13[0][0]'] \n",
            "                                                                                                  \n",
            " dropout_14 (Dropout)           (None, 64)           0           ['dense_17[0][0]']               \n",
            "                                                                                                  \n",
            " batch_normalization_14 (BatchN  (None, 64)          256         ['dropout_14[0][0]']             \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " concatenate_4 (Concatenate)    (None, 128)          0           ['batch_normalization_13[0][0]', \n",
            "                                                                  'batch_normalization_14[0][0]'] \n",
            "                                                                                                  \n",
            " dense_18 (Dense)               (None, 128)          16512       ['concatenate_4[0][0]']          \n",
            "                                                                                                  \n",
            " dropout_15 (Dropout)           (None, 128)          0           ['dense_18[0][0]']               \n",
            "                                                                                                  \n",
            " batch_normalization_15 (BatchN  (None, 128)         512         ['dropout_15[0][0]']             \n",
            " ormalization)                                                                                    \n",
            "                                                                                                  \n",
            " concatenate_5 (Concatenate)    (None, 256)          0           ['batch_normalization_12[0][0]', \n",
            "                                                                  'batch_normalization_15[0][0]'] \n",
            "                                                                                                  \n",
            " dense_19 (Dense)               (None, 50)           12850       ['concatenate_5[0][0]']          \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 49,842\n",
            "Trainable params: 49,074\n",
            "Non-trainable params: 768\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Anomaly detection\n",
        "anomaly_scores = []\n",
        "for inputs in val_data2_scaled:\n",
        "    inputs = tf.expand_dims(inputs, 0)  # Add batch dimension\n",
        "    outputs = model.predict(inputs)\n",
        "    loss = loss_fn(outputs, inputs)\n",
        "    anomaly_scores.append(loss.numpy())"
      ],
      "metadata": {
        "id": "PUaRzHAi0dYK",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cd0556c2-9eff-4eb9-8c67-dc0216324408"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 211ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 71ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 66ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 67ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 78ms/step\n",
            "1/1 [==============================] - 0s 68ms/step\n",
            "1/1 [==============================] - 0s 65ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 60ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 62ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.mean(anomaly_scores) + (np.std(anomaly_scores)/len(anomaly_scores)) # Adjust threshold based on your data and requirements\n",
        "threshold"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hHP8ZVt7t7aU",
        "outputId": "8dc33dff-32b8-4498-b342-a86cd5fa1a94"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.1448418786920364"
            ]
          },
          "metadata": {},
          "execution_count": 32
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        " np.mean(anomaly_scores) + np.std(anomaly_scores)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "z4-q6-SaOXrg",
        "outputId": "0dbd378e-60fe-43f1-af95-341f5872a201"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.2124656404193982"
            ]
          },
          "metadata": {},
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.hist(anomaly_scores)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "D5xOtaf_NfzR",
        "outputId": "38b3f759-da8b-4203-f4b7-300704bd9ca6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjAAAAGdCAYAAAAMm0nCAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAioklEQVR4nO3de3BU5eH/8U8uJKFINgTKbqKBREVB5KIgMV5RtgIyCGOs0qY0UgasJVqMF0jLpdZLAqWaQhEqo4AdkGpHUKGNYhBoNQQIYBUQwQaI4gaVZpdACYE8vz/6c7/dEpWEsyxP8n7NnBlzztmHZ585Ju852c1GGWOMAAAALBId6QkAAAA0FQEDAACsQ8AAAADrEDAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDqxkZ5AczQ0NOjAgQNq3769oqKiIj0dAABwGowxOnz4sFJTUxUdfWb3UKwMmAMHDigtLS3S0wAAAM1QVVWlCy644IzGsDJg2rdvL+k/C5CYmBjh2QAAgNMRCASUlpYW/Dl+JqwMmK9+bZSYmEjAAABgGSde/sGLeAEAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdZocMOvXr9fw4cOVmpqqqKgorVix4mvP/elPf6qoqCgVFxeH7D906JBycnKUmJiopKQkjR07VrW1tU2dCgAAaKWaHDBHjhxRnz59NHfu3G88b/ny5dqwYYNSU1NPOZaTk6Pt27dr9erVWrlypdavX6/x48c3dSoAAKCVavLfgRk6dKiGDh36jed8+umnuu+++/TGG29o2LBhIcd27typkpISbdq0Sf3795ckzZkzR7feeqtmzZrVaPAAAAD8N8dfA9PQ0KDRo0fr4YcfVs+ePU85XlZWpqSkpGC8SJLX61V0dLTKy8sbHbOurk6BQCBkAwAArZfjATNjxgzFxsbq/vvvb/S4z+dT586dQ/bFxsYqOTlZPp+v0ccUFhbK5XIFNz4HCQCA1s3RgKmoqNDvfvc7LVq0yNFPiS4oKJDf7w9uVVVVjo0NAADs42jA/O1vf9PBgwfVpUsXxcbGKjY2Vvv27dODDz6o9PR0SZLH49HBgwdDHnfixAkdOnRIHo+n0XHj4+ODn3vE5x8BAABHP8xx9OjR8nq9IfsGDx6s0aNHa8yYMZKkrKws1dTUqKKiQv369ZMkrVmzRg0NDcrMzHRyOgAAoIVqcsDU1tZqz549wa8rKyu1bds2JScnq0uXLurYsWPI+W3atJHH49Gll14qSerRo4eGDBmicePGaf78+aqvr1deXp5GjRrFO5AAAMBpaXLAbN68WTfddFPw6/z8fElSbm6uFi1adFpjLFmyRHl5eRo0aJCio6OVnZ2t2bNnN3UqYZM+eVWkp9Bke4uGfftJAAC0EE0OmIEDB8oYc9rn792795R9ycnJWrp0aVP/aQAAAEl8FhIAALAQAQMAAKxDwAAAAOsQMAAAwDoEDAAAsA4BAwAArEPAAAAA6xAwAADAOgQMAACwDgEDAACsQ8AAAADrEDAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDoEDAAAsA4BAwAArEPAAAAA6xAwAADAOgQMAACwDgEDAACsQ8AAAADrEDAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDoEDAAAsA4BAwAArEPAAAAA6xAwAADAOgQMAACwDgEDAACsQ8AAAADrEDAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDpNDpj169dr+PDhSk1NVVRUlFasWBE8Vl9fr0mTJqlXr15q166dUlNT9eMf/1gHDhwIGePQoUPKyclRYmKikpKSNHbsWNXW1p7xkwEAAK1DkwPmyJEj6tOnj+bOnXvKsaNHj2rLli2aOnWqtmzZoldeeUW7du3SbbfdFnJeTk6Otm/frtWrV2vlypVav369xo8f3/xnAQAAWpUoY4xp9oOjorR8+XKNHDnya8/ZtGmTBgwYoH379qlLly7auXOnLrvsMm3atEn9+/eXJJWUlOjWW2/VJ598otTU1G/9dwOBgFwul/x+vxITE5s7/a+VPnmV42OG296iYZGeAgAA38jJn99hfw2M3+9XVFSUkpKSJEllZWVKSkoKxoskeb1eRUdHq7y8vNEx6urqFAgEQjYAANB6hTVgjh07pkmTJukHP/hBsLR8Pp86d+4ccl5sbKySk5Pl8/kaHaewsFAulyu4paWlhXPaAADgHBe2gKmvr9edd94pY4zmzZt3RmMVFBTI7/cHt6qqKodmCQAAbBQbjkG/ipd9+/ZpzZo1Ib/n8ng8OnjwYMj5J06c0KFDh+TxeBodLz4+XvHx8eGYKgAAsJDjd2C+ipfdu3frrbfeUseOHUOOZ2VlqaamRhUVFcF9a9asUUNDgzIzM52eDgAAaIGafAemtrZWe/bsCX5dWVmpbdu2KTk5WSkpKbrjjju0ZcsWrVy5UidPngy+riU5OVlxcXHq0aOHhgwZonHjxmn+/Pmqr69XXl6eRo0adVrvQAIAAGhywGzevFk33XRT8Ov8/HxJUm5urn71q1/ptddekyT17ds35HFvv/22Bg4cKElasmSJ8vLyNGjQIEVHRys7O1uzZ89u5lMAAACtTZMDZuDAgfqmPx1zOn9WJjk5WUuXLm3qPw0AACCJz0ICAAAWImAAAIB1CBgAAGAdAgYAAFiHgAEAANYhYAAAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdQgYAABgHQIGAABYh4ABAADWIWAAAIB1CBgAAGAdAgYAAFiHgAEAANYhYAAAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdQgYAABgHQIGAABYh4ABAADWIWAAAIB1CBgAAGAdAgYAAFiHgAEAANYhYAAAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdQgYAABgHQIGAABYh4ABAADWIWAAAIB1CBgAAGAdAgYAAFinyQGzfv16DR8+XKmpqYqKitKKFStCjhtjNG3aNKWkpKht27byer3avXt3yDmHDh1STk6OEhMTlZSUpLFjx6q2tvaMnggAAGg9mhwwR44cUZ8+fTR37txGj8+cOVOzZ8/W/PnzVV5ernbt2mnw4ME6duxY8JycnBxt375dq1ev1sqVK7V+/XqNHz+++c8CAAC0KrFNfcDQoUM1dOjQRo8ZY1RcXKwpU6ZoxIgRkqQXXnhBbrdbK1as0KhRo7Rz506VlJRo06ZN6t+/vyRpzpw5uvXWWzVr1iylpqaewdMBAACtgaOvgamsrJTP55PX6w3uc7lcyszMVFlZmSSprKxMSUlJwXiRJK/Xq+joaJWXlzc6bl1dnQKBQMgGAABaL0cDxufzSZLcbnfIfrfbHTzm8/nUuXPnkOOxsbFKTk4OnvO/CgsL5XK5gltaWpqT0wYAAJax4l1IBQUF8vv9wa2qqirSUwIAABHkaMB4PB5JUnV1dcj+6urq4DGPx6ODBw+GHD9x4oQOHToUPOd/xcfHKzExMWQDAACtl6MBk5GRIY/Ho9LS0uC+QCCg8vJyZWVlSZKysrJUU1OjioqK4Dlr1qxRQ0ODMjMznZwOAABooZr8LqTa2lrt2bMn+HVlZaW2bdum5ORkdenSRRMnTtTjjz+ubt26KSMjQ1OnTlVqaqpGjhwpSerRo4eGDBmicePGaf78+aqvr1deXp5GjRrFO5AAAMBpaXLAbN68WTfddFPw6/z8fElSbm6uFi1apEceeURHjhzR+PHjVVNTo+uuu04lJSVKSEgIPmbJkiXKy8vToEGDFB0drezsbM2ePduBpwMAAFqDKGOMifQkmioQCMjlcsnv94fl9TDpk1c5Pma47S0aFukpAADwjZz8+W3Fu5AAAAD+GwEDAACsQ8AAAADrEDAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDoEDAAAsA4BAwAArEPAAAAA6xAwAADAOgQMAACwDgEDAACsQ8AAAADrEDAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDoEDAAAsE5spCcAZ6RPXhXpKTTZ3qJhkZ4CAMBS3IEBAADWIWAAAIB1CBgAAGAdAgYAAFiHgAEAANYhYAAAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdQgYAABgHQIGAABYh4ABAADWIWAAAIB1CBgAAGAdAgYAAFiHgAEAANaJjfQE0HqlT14V6Sk02d6iYZGeAgBAYbgDc/LkSU2dOlUZGRlq27atLrroIj322GMyxgTPMcZo2rRpSklJUdu2beX1erV7926npwIAAFooxwNmxowZmjdvnn7/+99r586dmjFjhmbOnKk5c+YEz5k5c6Zmz56t+fPnq7y8XO3atdPgwYN17Ngxp6cDAABaIMd/hfTuu+9qxIgRGjbsP7fa09PT9eKLL2rjxo2S/nP3pbi4WFOmTNGIESMkSS+88ILcbrdWrFihUaNGOT0lAADQwjh+B+aaa65RaWmpPvroI0nSe++9p7///e8aOnSoJKmyslI+n09erzf4GJfLpczMTJWVlTU6Zl1dnQKBQMgGAABaL8fvwEyePFmBQEDdu3dXTEyMTp48qSeeeEI5OTmSJJ/PJ0lyu90hj3O73cFj/6uwsFCPPvqo01MFAACWcvwOzEsvvaQlS5Zo6dKl2rJlixYvXqxZs2Zp8eLFzR6zoKBAfr8/uFVVVTk4YwAAYBvH78A8/PDDmjx5cvC1LL169dK+fftUWFio3NxceTweSVJ1dbVSUlKCj6uurlbfvn0bHTM+Pl7x8fFOTxUAAFjK8TswR48eVXR06LAxMTFqaGiQJGVkZMjj8ai0tDR4PBAIqLy8XFlZWU5PBwAAtECO34EZPny4nnjiCXXp0kU9e/bU1q1b9dRTT+knP/mJJCkqKkoTJ07U448/rm7duikjI0NTp05VamqqRo4c6fR0AABAC+R4wMyZM0dTp07Vz372Mx08eFCpqam65557NG3atOA5jzzyiI4cOaLx48erpqZG1113nUpKSpSQkOD0dAAAQAsUZf77T+RaIhAIyOVyye/3KzEx0fHxbfwT9zg7+CgBAGg+J39+82GOAADAOgQMAACwDgEDAACsQ8AAAADrEDAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDoEDAAAsA4BAwAArEPAAAAA6xAwAADAOgQMAACwDgEDAACsQ8AAAADrEDAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDoEDAAAsA4BAwAArEPAAAAA6xAwAADAOgQMAACwTmykJwDYJH3yqkhPocn2Fg2L9BQAwHHcgQEAANYhYAAAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdQgYAABgHQIGAABYh4ABAADWIWAAAIB1CBgAAGAdAgYAAFiHgAEAANYJS8B8+umn+tGPfqSOHTuqbdu26tWrlzZv3hw8bozRtGnTlJKSorZt28rr9Wr37t3hmAoAAGiBHA+Yf/3rX7r22mvVpk0b/fWvf9WOHTv029/+Vh06dAieM3PmTM2ePVvz589XeXm52rVrp8GDB+vYsWNOTwcAALRAsU4POGPGDKWlpWnhwoXBfRkZGcH/NsaouLhYU6ZM0YgRIyRJL7zwgtxut1asWKFRo0Y5PSUAANDCOH4H5rXXXlP//v31/e9/X507d9YVV1yhBQsWBI9XVlbK5/PJ6/UG97lcLmVmZqqsrKzRMevq6hQIBEI2AADQejkeMP/85z81b948devWTW+88Ybuvfde3X///Vq8eLEkyefzSZLcbnfI49xud/DY/yosLJTL5QpuaWlpTk8bAABYxPGAaWho0JVXXqknn3xSV1xxhcaPH69x48Zp/vz5zR6zoKBAfr8/uFVVVTk4YwAAYBvHAyYlJUWXXXZZyL4ePXpo//79kiSPxyNJqq6uDjmnuro6eOx/xcfHKzExMWQDAACtl+MBc+2112rXrl0h+z766CN17dpV0n9e0OvxeFRaWho8HggEVF5erqysLKenAwAAWiDH34X0wAMP6JprrtGTTz6pO++8Uxs3btSzzz6rZ599VpIUFRWliRMn6vHHH1e3bt2UkZGhqVOnKjU1VSNHjnR6OgAAoAVyPGCuuuoqLV++XAUFBfr1r3+tjIwMFRcXKycnJ3jOI488oiNHjmj8+PGqqanRddddp5KSEiUkJDg9HQAA0AJFGWNMpCfRVIFAQC6XS36/Pyyvh0mfvMrxMYFI2Vs0LNJTAABJzv785rOQAACAdQgYAABgHQIGAABYh4ABAADWIWAAAIB1CBgAAGAdAgYAAFiHgAEAANYhYAAAgHUIGAAAYB0CBgAAWMfxD3MEcG6x8bO9+PwmAN+GOzAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDoEDAAAsA4BAwAArEPAAAAA6xAwAADAOgQMAACwDgEDAACsQ8AAAADr8GGOAM45fAAlgG/DHRgAAGAdAgYAAFiHgAEAANYhYAAAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdQgYAABgHf4SLwA4gL8eDJxd3IEBAADWIWAAAIB1CBgAAGAdAgYAAFiHgAEAANYJe8AUFRUpKipKEydODO47duyYJkyYoI4dO+q8885Tdna2qqurwz0VAADQQoQ1YDZt2qQ//OEP6t27d8j+Bx54QK+//rpefvllrVu3TgcOHNDtt98ezqkAAIAWJGwBU1tbq5ycHC1YsEAdOnQI7vf7/Xruuef01FNP6eabb1a/fv20cOFCvfvuu9qwYUO4pgMAAFqQsAXMhAkTNGzYMHm93pD9FRUVqq+vD9nfvXt3denSRWVlZY2OVVdXp0AgELIBAIDWKyx/iXfZsmXasmWLNm3adMoxn8+nuLg4JSUlhex3u93y+XyNjldYWKhHH300HFMFAAAWcvwOTFVVlX7+859ryZIlSkhIcGTMgoIC+f3+4FZVVeXIuAAAwE6OB0xFRYUOHjyoK6+8UrGxsYqNjdW6des0e/ZsxcbGyu126/jx46qpqQl5XHV1tTweT6NjxsfHKzExMWQDAACtl+O/Qho0aJDef//9kH1jxoxR9+7dNWnSJKWlpalNmzYqLS1Vdna2JGnXrl3av3+/srKynJ4OAABogRwPmPbt2+vyyy8P2deuXTt17NgxuH/s2LHKz89XcnKyEhMTdd999ykrK0tXX32109MBAAAtUFhexPttnn76aUVHRys7O1t1dXUaPHiwnnnmmUhMBQAAWCjKGGMiPYmmCgQCcrlc8vv9YXk9TPrkVY6PCQDnmr1FwyI9BbQyTv785rOQAACAdQgYAABgHQIGAABYh4ABAADWIWAAAIB1CBgAAGAdAgYAAFiHgAEAANYhYAAAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdQgYAABgndhITwAAEBnpk1dFegpNtrdoWKSngHMEd2AAAIB1CBgAAGAdAgYAAFiHgAEAANYhYAAAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdQgYAABgHQIGAABYh4ABAADWIWAAAIB1CBgAAGAdAgYAAFiHgAEAANYhYAAAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdQgYAABgHQIGAABYh4ABAADWIWAAAIB1HA+YwsJCXXXVVWrfvr06d+6skSNHateuXSHnHDt2TBMmTFDHjh113nnnKTs7W9XV1U5PBQAAtFCOB8y6des0YcIEbdiwQatXr1Z9fb1uueUWHTlyJHjOAw88oNdff10vv/yy1q1bpwMHDuj22293eioAAKCFinV6wJKSkpCvFy1apM6dO6uiokI33HCD/H6/nnvuOS1dulQ333yzJGnhwoXq0aOHNmzYoKuvvtrpKQEAgBYm7K+B8fv9kqTk5GRJUkVFherr6+X1eoPndO/eXV26dFFZWVmjY9TV1SkQCIRsAACg9QprwDQ0NGjixIm69tprdfnll0uSfD6f4uLilJSUFHKu2+2Wz+drdJzCwkK5XK7glpaWFs5pAwCAc1xYA2bChAn64IMPtGzZsjMap6CgQH6/P7hVVVU5NEMAAGAjx18D85W8vDytXLlS69ev1wUXXBDc7/F4dPz4cdXU1ITchamurpbH42l0rPj4eMXHx4drqgAAwDKO34ExxigvL0/Lly/XmjVrlJGREXK8X79+atOmjUpLS4P7du3apf379ysrK8vp6QAAgBbI8TswEyZM0NKlS/Xqq6+qffv2wde1uFwutW3bVi6XS2PHjlV+fr6Sk5OVmJio++67T1lZWbwDCQAAnBbHA2bevHmSpIEDB4bsX7hwoe6++25J0tNPP63o6GhlZ2errq5OgwcP1jPPPOP0VAAAQAvleMAYY771nISEBM2dO1dz5851+p8HAACtAJ+FBAAArEPAAAAA6xAwAADAOgQMAACwDgEDAACsQ8AAAADrEDAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDoEDAAAsA4BAwAArEPAAAAA6xAwAADAOgQMAACwTmykJwAAwOlKn7wq0lNosr1FwyI9hRaJOzAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDoEDAAAsA4BAwAArEPAAAAA6xAwAADAOgQMAACwDgEDAACsQ8AAAADrEDAAAMA6BAwAALAOAQMAAKxDwAAAAOsQMAAAwDoEDAAAsA4BAwAArEPAAAAA68RGegIAALRk6ZNXRXoKzbK3aFikp/CNuAMDAACsQ8AAAADrRDRg5s6dq/T0dCUkJCgzM1MbN26M5HQAAIAlIhYwf/rTn5Sfn6/p06dry5Yt6tOnjwYPHqyDBw9GakoAAMASEQuYp556SuPGjdOYMWN02WWXaf78+frOd76j559/PlJTAgAAlojIu5COHz+uiooKFRQUBPdFR0fL6/WqrKzslPPr6upUV1cX/Nrv90uSAoFAWObXUHc0LOMCAGCLcPyM/WpMY8wZjxWRgPniiy908uRJud3ukP1ut1sffvjhKecXFhbq0UcfPWV/Wlpa2OYIAEBr5ioO39iHDx+Wy+U6ozGs+DswBQUFys/PD37d0NCgQ4cOqWPHjoqKimrSWIFAQGlpaaqqqlJiYqLTU23RWLvmY+2aj7VrPtau+Vi75vumtTPG6PDhw0pNTT3jfyciAdOpUyfFxMSouro6ZH91dbU8Hs8p58fHxys+Pj5kX1JS0hnNITExkYuymVi75mPtmo+1az7WrvlYu+b7urU70zsvX4nIi3jj4uLUr18/lZaWBvc1NDSotLRUWVlZkZgSAACwSMR+hZSfn6/c3Fz1799fAwYMUHFxsY4cOaIxY8ZEakoAAMASEQuYu+66S59//rmmTZsmn8+nvn37qqSk5JQX9jotPj5e06dPP+VXUvh2rF3zsXbNx9o1H2vXfKxd852ttYsyTryXCQAA4Czis5AAAIB1CBgAAGAdAgYAAFiHgAEAANaxPmDmzp2r9PR0JSQkKDMzUxs3bvzG819++WV1795dCQkJ6tWrl/7yl7+EHL/77rsVFRUVsg0ZMiScTyFimrJ227dvV3Z2ttLT0xUVFaXi4uIzHtNmTq/dr371q1Ouu+7du4fxGUROU9ZuwYIFuv7669WhQwd16NBBXq/3lPONMZo2bZpSUlLUtm1beb1e7d69O9xPI2KcXj++5zXulVdeUf/+/ZWUlKR27dqpb9+++uMf/xhyTmu69pxeO0euO2OxZcuWmbi4OPP888+b7du3m3HjxpmkpCRTXV3d6PnvvPOOiYmJMTNnzjQ7duwwU6ZMMW3atDHvv/9+8Jzc3FwzZMgQ89lnnwW3Q4cOna2ndNY0de02btxoHnroIfPiiy8aj8djnn766TMe01bhWLvp06ebnj17hlx3n3/+eZifydnX1LX74Q9/aObOnWu2bt1qdu7cae6++27jcrnMJ598EjynqKjIuFwus2LFCvPee++Z2267zWRkZJh///vfZ+tpnTXhWD++5zW+dm+//bZ55ZVXzI4dO8yePXtMcXGxiYmJMSUlJcFzWsu1F461c+K6szpgBgwYYCZMmBD8+uTJkyY1NdUUFhY2ev6dd95phg0bFrIvMzPT3HPPPcGvc3NzzYgRI8Iy33NJU9fuv3Xt2rXRH8JnMqZNwrF206dPN3369HFwluemM71GTpw4Ydq3b28WL15sjDGmoaHBeDwe85vf/CZ4Tk1NjYmPjzcvvviis5M/Bzi9fsbwPa8p35+uuOIKM2XKFGNM67r2nF47Y5y57qz9FdLx48dVUVEhr9cb3BcdHS2v16uysrJGH1NWVhZyviQNHjz4lPPXrl2rzp0769JLL9W9996rL7/80vknEEHNWbtIjHkuCufz3L17t1JTU3XhhRcqJydH+/fvP9PpnlOcWLujR4+qvr5eycnJkqTKykr5fL6QMV0ulzIzM1vUdSeFZ/2+wve8b2aMUWlpqXbt2qUbbrhBUuu59sKxdl850+vOik+jbswXX3yhkydPnvKXe91utz788MNGH+Pz+Ro93+fzBb8eMmSIbr/9dmVkZOjjjz/WL37xCw0dOlRlZWWKiYlx/olEQHPWLhJjnovC9TwzMzO1aNEiXXrppfrss8/06KOP6vrrr9cHH3yg9u3bn+m0zwlOrN2kSZOUmpoa/Gb61f+73/b/dUsQjvWT+J73TWvn9/t1/vnnq66uTjExMXrmmWf0ve99T1LrufbCsXaSM9edtQETLqNGjQr+d69evdS7d29ddNFFWrt2rQYNGhTBmaElGzp0aPC/e/furczMTHXt2lUvvfSSxo4dG8GZnTuKioq0bNkyrV27VgkJCZGejnW+bv34nvf12rdvr23btqm2tlalpaXKz8/XhRdeqIEDB0Z6aue8b1s7J647a3+F1KlTJ8XExKi6ujpkf3V1tTweT6OP8Xg8TTpfki688EJ16tRJe/bsOfNJnyOas3aRGPNcdLaeZ1JSki655BKuu/9v1qxZKioq0ptvvqnevXsH93/1uJZ+3UnhWb/G8D3v/0RHR+viiy9W37599eCDD+qOO+5QYWGhpNZz7YVj7RrTnOvO2oCJi4tTv379VFpaGtzX0NCg0tJSZWVlNfqYrKyskPMlafXq1V97viR98skn+vLLL5WSkuLMxM8BzVm7SIx5Ljpbz7O2tlYff/wx152kmTNn6rHHHlNJSYn69+8fciwjI0MejydkzEAgoPLy8hZ13UnhWb/G8D3v6zU0NKiurk5S67n2wrF2jWnWdXdGLwGOsGXLlpn4+HizaNEis2PHDjN+/HiTlJRkfD6fMcaY0aNHm8mTJwfPf+edd0xsbKyZNWuW2blzp5k+fXrI26gPHz5sHnroIVNWVmYqKyvNW2+9Za688krTrVs3c+zYsYg8x3Bp6trV1dWZrVu3mq1bt5qUlBTz0EMPma1bt5rdu3ef9pgtRTjW7sEHHzRr1641lZWV5p133jFer9d06tTJHDx48Kw/v3Bq6toVFRWZuLg48+c//znk7ZaHDx8OOScpKcm8+uqr5h//+IcZMWJEi3wrqzHOrx/f875+7Z588knz5ptvmo8//tjs2LHDzJo1y8TGxpoFCxYEz2kt157Ta+fUdWd1wBhjzJw5c0yXLl1MXFycGTBggNmwYUPw2I033mhyc3NDzn/ppZfMJZdcYuLi4kzPnj3NqlWrgseOHj1qbrnlFvPd737XtGnTxnTt2tWMGzeuxf0A/kpT1q6ystJIOmW78cYbT3vMlsTptbvrrrtMSkqKiYuLM+eff7656667zJ49e87iMzp7mrJ2Xbt2bXTtpk+fHjynoaHBTJ061bjdbhMfH28GDRpkdu3adRaf0dnl5PrxPe/r1+6Xv/ylufjii01CQoLp0KGDycrKMsuWLQsZrzVde06unVPXXZQxxpz+/RoAAIDIs/Y1MAAAoPUiYAAAgHUIGAAAYB0CBgAAWIeAAQAA1iFgAACAdQgYAABgHQIGAABYh4ABAADWIWAAAIB1CBgAAGAdAgYAAFjn/wEOhD5XFTlcDQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Anomaly detection\n",
        "anomaly_scores = []\n",
        "for inputs in test_data_scaled:\n",
        "    inputs = tf.expand_dims(inputs, 0)  # Add batch dimension\n",
        "    outputs = model.predict(inputs)\n",
        "    loss = loss_fn(outputs, inputs)\n",
        "    anomaly_scores.append(loss.numpy())"
      ],
      "metadata": {
        "id": "XLIptWmSuAbe",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2e9d2e27-834b-4adf-e102-79eb91d85406"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n",
            "1/1 [==============================] - 0s 81ms/step\n",
            "1/1 [==============================] - 0s 86ms/step\n",
            "1/1 [==============================] - 0s 99ms/step\n",
            "1/1 [==============================] - 0s 117ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Thresholding and labeling anomalies\n",
        "anomalies = [1 if score > threshold else 0 for score in anomaly_scores]"
      ],
      "metadata": {
        "id": "oK0uc_Hk0_8c"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(gt, anomalies, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gVWyteZ1N245",
        "outputId": "d67d199e-fa75-4dd9-d680-8d52a1723dd6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.66      0.87      0.75        71\n",
            "      attack       0.65      0.35      0.45        49\n",
            "\n",
            "    accuracy                           0.66       120\n",
            "   macro avg       0.66      0.61      0.60       120\n",
            "weighted avg       0.66      0.66      0.63       120\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Isolation Forest"
      ],
      "metadata": {
        "id": "PTHg4m_6YZcO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.ensemble import IsolationForest\n",
        "from sklearn.metrics import precision_score, recall_score, f1_score\n",
        "\n",
        "# Train the Isolation Forest model\n",
        "isolation_forest = IsolationForest(contamination=0.0007)  # Adjust the contamination parameter as needed\n",
        "isolation_forest.fit(train_data_scaled)\n",
        "\n",
        "# Classify the test data\n",
        "y_pred = isolation_forest.predict(test_data_scaled)\n",
        "y_pred = np.where(y_pred == -1, 1, 0)  # Convert -1 labels to 1 for anomalies\n",
        "\n",
        "# Evaluate the performance on the test data\n",
        "precision = precision_score(gt, y_pred)\n",
        "recall = recall_score(gt, y_pred)\n",
        "f1 = f1_score(gt, y_pred)\n",
        "\n",
        "# Print the evaluation metrics\n",
        "print(\"Precision:\", precision)\n",
        "print(\"Recall:\", recall)\n",
        "print(\"F1-Score:\", f1)\n"
      ],
      "metadata": {
        "id": "QYQNLoJ4vUgA",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b788da76-cbef-43db-9c14-f5e61089ee07"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Precision: 0.5465116279069767\n",
            "Recall: 0.9591836734693877\n",
            "F1-Score: 0.6962962962962962\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Au4EZVKSlsR5",
        "outputId": "73e477a3-d1d3-400e-e7fb-d53a42fdce3e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0,\n",
              "       0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1,\n",
              "       1, 0, 0, 0, 0, 0, 0, 1, 0, 1])"
            ]
          },
          "metadata": {},
          "execution_count": 471
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(gt, y_pred, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "id": "E3zOd74EZjjR",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3381f15a-3349-4054-85b8-5a00ce0a3e60"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.94      0.45      0.61        71\n",
            "      attack       0.55      0.96      0.70        49\n",
            "\n",
            "    accuracy                           0.66       120\n",
            "   macro avg       0.74      0.70      0.65       120\n",
            "weighted avg       0.78      0.66      0.64       120\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "##Local Outlier Factor"
      ],
      "metadata": {
        "id": "TnGAab2_CAXa"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.neighbors import LocalOutlierFactor\n",
        "from sklearn.metrics import precision_score, recall_score, f1_score\n",
        "\n",
        "# Train the LOF model\n",
        "lof_model = LocalOutlierFactor(n_neighbors=5, contamination=0.001, novelty=True)  # Adjust parameters as needed\n",
        "lof_model.fit(train_data_scaled)\n",
        "\n",
        "# Classify the test data\n",
        "y_pred = lof_model.predict(test_data_scaled)\n",
        "y_pred = np.where(y_pred == -1, 1, 0)  # Convert -1 labels to 1 for anomalies\n",
        "\n",
        "# Evaluate the performance on the test data\n",
        "precision = precision_score(gt, y_pred)\n",
        "recall = recall_score(gt, y_pred)\n",
        "f1 = f1_score(gt, y_pred)\n",
        "\n",
        "# Print the evaluation metrics\n",
        "print(\"Precision:\", precision)\n",
        "print(\"Recall:\", recall)\n",
        "print(\"F1-Score:\", f1)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "veFWO7PXB_43",
        "outputId": "1a41399b-ff15-4850-8d0d-fb0ea57fe37b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Precision: 0.4083333333333333\n",
            "Recall: 1.0\n",
            "F1-Score: 0.5798816568047338\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2Gnu9Y8mDYs-",
        "outputId": "55ff5275-89e8-4b49-837e-2534db740377"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1])"
            ]
          },
          "metadata": {},
          "execution_count": 56
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "##OCSVM"
      ],
      "metadata": {
        "id": "EYdcVGiSEJdj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.svm import OneClassSVM\n",
        "\n",
        "# Train the One-Class SVM\n",
        "svm_model = OneClassSVM(kernel='rbf', nu=0.0005)  # Adjust parameters as needed\n",
        "svm_model.fit(train_data_scaled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 75
        },
        "id": "aTePTCFdg61V",
        "outputId": "e27a1156-839d-4fd6-d346-c6398ac7efb6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "OneClassSVM(nu=0.0005)"
            ],
            "text/html": [
              "<style>#sk-container-id-8 {color: black;background-color: white;}#sk-container-id-8 pre{padding: 0;}#sk-container-id-8 div.sk-toggleable {background-color: white;}#sk-container-id-8 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-8 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-8 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-8 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-8 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-8 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-8 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-8 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-8 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-8 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-8 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-8 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-8 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-8 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-8 div.sk-item {position: relative;z-index: 1;}#sk-container-id-8 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-8 div.sk-item::before, #sk-container-id-8 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-8 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-8 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-8 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-8 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-8 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-8 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-8 div.sk-label-container {text-align: center;}#sk-container-id-8 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-8 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-8\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>OneClassSVM(nu=0.0005)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-8\" type=\"checkbox\" checked><label for=\"sk-estimator-id-8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">OneClassSVM</label><div class=\"sk-toggleable__content\"><pre>OneClassSVM(nu=0.0005)</pre></div></div></div></div></div>"
            ]
          },
          "metadata": {},
          "execution_count": 64
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "svm_model.fit_predict(val_data1_scaled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KjUt1i9vhtuJ",
        "outputId": "e17e8271-58a0-4b42-db2f-58e27a004870"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1,  1,  1,  1,\n",
              "        1,  1,  1,  1, -1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1, -1,  1, -1,  1, -1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1])"
            ]
          },
          "metadata": {},
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "svm_model.fit_predict(val_data2_scaled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Zq1-iokrkMwU",
        "outputId": "1ffa2c8c-3b86-4aba-bf7c-cf0b3113be0a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1, -1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1, -1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1, -1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1, -1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1, -1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1, -1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1, -1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,\n",
              "        1,  1,  1,  1,  1,  1])"
            ]
          },
          "metadata": {},
          "execution_count": 66
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred = svm_model.predict(test_data_scaled)"
      ],
      "metadata": {
        "id": "5eSN61cqg_WY"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "binary_labels = np.where(pred == -1, 1, 0)"
      ],
      "metadata": {
        "id": "1MJk5BR6h_Yd"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "binary_labels"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ufjHVkNhifpx",
        "outputId": "be5a9392-df33-4c8e-b845-12e275b3fde2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0,\n",
              "       0, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0,\n",
              "       1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 0, 1, 1, 0, 1])"
            ]
          },
          "metadata": {},
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(gt, binary_labels, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FssSmkptimvX",
        "outputId": "dffe56c7-a886-4cac-d818-b9f76f2ece9a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.88      0.32      0.47        71\n",
            "      attack       0.49      0.94      0.64        49\n",
            "\n",
            "    accuracy                           0.57       120\n",
            "   macro avg       0.69      0.63      0.56       120\n",
            "weighted avg       0.72      0.57      0.54       120\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.svm import OneClassSVM\n",
        "\n",
        "# Train the One-Class SVM\n",
        "svm_model = OneClassSVM(kernel='linear', nu=0.9)  # Adjust parameters as needed\n",
        "svm_model.fit(train_data_scaled)\n",
        "\n",
        "# Obtain the decision function values for test data\n",
        "decision_values = svm_model.decision_function(test_data_scaled)\n",
        "\n",
        "# Plot the decision function values\n",
        "plt.figure(figsize=(10, 6))\n",
        "plt.plot(decision_values, 'ro')\n",
        "plt.xlabel('Test Data Point')\n",
        "plt.ylabel('Decision Function Value')\n",
        "plt.title('One-Class SVM Decision Function Values (kernel=rbf)')\n",
        "plt.grid(True)\n",
        "plt.show()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 564
        },
        "id": "-j1TnS2cc7X2",
        "outputId": "2710a28b-0503-4127-b676-aefe84985f00"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Obtain the decision function values for test data\n",
        "decision_values = svm_model.decision_function(val_data2_scaled)\n",
        "\n",
        "# Plot the decision function values\n",
        "plt.figure(figsize=(10, 6))\n",
        "plt.plot(decision_values, 'ro')\n",
        "plt.xlabel('Validation Data Point')\n",
        "plt.ylabel('Decision Function Value')\n",
        "plt.title('One-Class SVM Decision Function Values (kernel=rbf)')\n",
        "plt.grid(True)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 564
        },
        "id": "wevRBXxGdJc1",
        "outputId": "cc115846-d98b-4788-e226-0a715b1717e7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.hist(svm_model.decision_function(val_data2_scaled))\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 433
        },
        "id": "Nbi_hjwVPWuK",
        "outputId": "28c19a28-c3bb-49b2-8a37-aabb01060b71"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAigAAAGgCAYAAACABpytAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAkUUlEQVR4nO3dfXBU1eH/8c+GkE0C7IaEZpe0oaQtCioFFIlB2vEhNVqqoGkVTS0iQq2hLcYKZEbwi4pBVKRQnnSUhxmRSmdEFI2DQUVrCBDAJxDpFDRVNthCdgUlBHJ+f7Tcn0uCJHLDng3v18yddu89ezknm5J3L3c3HmOMEQAAgEUSYj0BAACA4xEoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDqtDpR169bp6quvVlZWljwej1auXOkca2ho0MSJE9W3b1916tRJWVlZ+s1vfqPPPvss6hz79u1TUVGRfD6f0tLSNHr0aB04cOCUFwMAANqHxNY+4eDBg+rXr59uvfVWXXfddVHHvvzyS23evFmTJ09Wv379tH//fv3xj3/UNddco02bNjnjioqKtGfPHq1Zs0YNDQ0aNWqUxo4dq2XLlrVoDo2Njfrss8/UpUsXeTye1i4BAADEgDFGX3zxhbKyspSQcJJrJOYUSDLPPffcN47ZsGGDkWQ+/vhjY4wx27ZtM5LMxo0bnTEvv/yy8Xg85tNPP23Rn1tTU2MksbGxsbGxscXhVlNTc9Kf9a2+gtJa4XBYHo9HaWlpkqTKykqlpaVp4MCBzpj8/HwlJCSoqqpK1157bZNz1NfXq76+3nls/vcLmGtqauTz+dp2AQAAwBWRSETZ2dnq0qXLSce2aaAcOnRIEydO1I033uiERCgUUmZmZvQkEhOVnp6uUCjU7HnKyso0derUJvt9Ph+BAgBAnGnJ7Rlt9i6ehoYGXX/99TLGaP78+ad0rtLSUoXDYWerqalxaZYAAMBGbXIF5VicfPzxx1q7dm3UVY5gMKi9e/dGjT9y5Ij27dunYDDY7Pm8Xq+8Xm9bTBUAAFjI9Ssox+Jk586devXVV5WRkRF1PC8vT3V1daqurnb2rV27Vo2NjcrNzXV7OgAAIA61+grKgQMH9I9//MN5vGvXLm3dulXp6enq3r27fvnLX2rz5s168cUXdfToUee+kvT0dCUlJalPnz668sorNWbMGC1YsEANDQ0aN26cRowYoaysLPdWBgAA4pbHHHtLTAu9/vrruvTSS5vsHzlypP7v//5POTk5zT7vtdde0yWXXCLpvx/UNm7cOL3wwgtKSEhQYWGhZs+erc6dO7doDpFIRH6/X+FwmJtkAQCIE635+d3qQLEBgQIAQPxpzc9vfhcPAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDpt8ssCAdij56TVsZ5Cq+2ePjTWUwAQY1xBAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWaXWgrFu3TldffbWysrLk8Xi0cuXKqOPGGE2ZMkXdu3dXSkqK8vPztXPnzqgx+/btU1FRkXw+n9LS0jR69GgdOHDglBYCAADaj1YHysGDB9WvXz/NnTu32eMzZszQ7NmztWDBAlVVValTp04qKCjQoUOHnDFFRUX64IMPtGbNGr344otat26dxo4d++1XAQAA2pXE1j7hqquu0lVXXdXsMWOMZs2apXvuuUfDhg2TJC1dulSBQEArV67UiBEjtH37dpWXl2vjxo0aOHCgJGnOnDn6+c9/rkceeURZWVmnsBwAANAeuHoPyq5duxQKhZSfn+/s8/v9ys3NVWVlpSSpsrJSaWlpTpxIUn5+vhISElRVVdXseevr6xWJRKI2AADQfrkaKKFQSJIUCASi9gcCAedYKBRSZmZm1PHExESlp6c7Y45XVlYmv9/vbNnZ2W5OGwAAWCYu3sVTWlqqcDjsbDU1NbGeEgAAaEOuBkowGJQk1dbWRu2vra11jgWDQe3duzfq+JEjR7Rv3z5nzPG8Xq98Pl/UBgAA2i9XAyUnJ0fBYFAVFRXOvkgkoqqqKuXl5UmS8vLyVFdXp+rqamfM2rVr1djYqNzcXDenAwAA4lSr38Vz4MAB/eMf/3Ae79q1S1u3blV6erp69Oih8ePH64EHHlCvXr2Uk5OjyZMnKysrS8OHD5ck9enTR1deeaXGjBmjBQsWqKGhQePGjdOIESN4Bw8AAJD0LQJl06ZNuvTSS53HJSUlkqSRI0dq8eLFmjBhgg4ePKixY8eqrq5OQ4YMUXl5uZKTk53nPP300xo3bpwuv/xyJSQkqLCwULNnz3ZhOQAAoD3wGGNMrCfRWpFIRH6/X+FwmPtRgJPoOWl1rKfQarunD431FAC0gdb8/I6Ld/EAAIAzC4ECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrJMZ6AnBHz0mrYz2FVts9fWispwAAsBRXUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADW4YPaEDN8uBwA4ES4ggIAAKzjeqAcPXpUkydPVk5OjlJSUvTDH/5Q999/v4wxzhhjjKZMmaLu3bsrJSVF+fn52rlzp9tTAQAAccr1QHnooYc0f/58/eUvf9H27dv10EMPacaMGZozZ44zZsaMGZo9e7YWLFigqqoqderUSQUFBTp06JDb0wEAAHHI9XtQ3n77bQ0bNkxDh/733+p79uypZ555Rhs2bJD036sns2bN0j333KNhw4ZJkpYuXapAIKCVK1dqxIgRbk8JAADEGdevoAwePFgVFRX66KOPJEnvvPOO3nrrLV111VWSpF27dikUCik/P995jt/vV25uriorK5s9Z319vSKRSNQGAADaL9evoEyaNEmRSES9e/dWhw4ddPToUU2bNk1FRUWSpFAoJEkKBAJRzwsEAs6x45WVlWnq1KluTxUAAFjK9Ssozz77rJ5++mktW7ZMmzdv1pIlS/TII49oyZIl3/qcpaWlCofDzlZTU+PijAEAgG1cv4Jy9913a9KkSc69JH379tXHH3+ssrIyjRw5UsFgUJJUW1ur7t27O8+rra1V//79mz2n1+uV1+t1e6oAAMBSrl9B+fLLL5WQEH3aDh06qLGxUZKUk5OjYDCoiooK53gkElFVVZXy8vLcng4AAIhDrl9BufrqqzVt2jT16NFD5557rrZs2aKZM2fq1ltvlSR5PB6NHz9eDzzwgHr16qWcnBxNnjxZWVlZGj58uNvTAQAAccj1QJkzZ44mT56sO+64Q3v37lVWVpZ++9vfasqUKc6YCRMm6ODBgxo7dqzq6uo0ZMgQlZeXKzk52e3pAACAOOQxX/+I1zgRiUTk9/sVDofl8/liPR0rxOPvtYlH8fi7eOLxeyMev84ATq41P7/5XTwAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKyTGOsJAPGk56TVsZ7CGSEev867pw+N9RSAdoUrKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrJMZ6AgDQHvSctDrWU2i13dOHxnoKwAlxBQUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANZpk0D59NNP9etf/1oZGRlKSUlR3759tWnTJue4MUZTpkxR9+7dlZKSovz8fO3cubMtpgIAAOKQ64Gyf/9+XXzxxerYsaNefvllbdu2TY8++qi6du3qjJkxY4Zmz56tBQsWqKqqSp06dVJBQYEOHTrk9nQAAEAcSnT7hA899JCys7O1aNEiZ19OTo7z340xmjVrlu655x4NGzZMkrR06VIFAgGtXLlSI0aMcHtKAAAgzrh+BWXVqlUaOHCgfvWrXykzM1MDBgzQE0884RzftWuXQqGQ8vPznX1+v1+5ubmqrKxs9pz19fWKRCJRGwAAaL9cD5R//vOfmj9/vnr16qVXXnlFv/vd7/SHP/xBS5YskSSFQiFJUiAQiHpeIBBwjh2vrKxMfr/f2bKzs92eNgAAsIjrgdLY2Kjzzz9fDz74oAYMGKCxY8dqzJgxWrBgwbc+Z2lpqcLhsLPV1NS4OGMAAGAb1wOle/fuOuecc6L29enTR5988okkKRgMSpJqa2ujxtTW1jrHjuf1euXz+aI2AADQfrkeKBdffLF27NgRte+jjz7S97//fUn/vWE2GAyqoqLCOR6JRFRVVaW8vDy3pwMAAOKQ6+/iufPOOzV48GA9+OCDuv7667VhwwY9/vjjevzxxyVJHo9H48eP1wMPPKBevXopJydHkydPVlZWloYPH+72dAAAQBxyPVAuvPBCPffccyotLdV9992nnJwczZo1S0VFRc6YCRMm6ODBgxo7dqzq6uo0ZMgQlZeXKzk52e3pAACAOOQxxphYT6K1IpGI/H6/wuEw96P8T89Jq2M9BQBxZvf0obGeAs4wrfn5ze/iAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAddo8UKZPny6Px6Px48c7+w4dOqTi4mJlZGSoc+fOKiwsVG1tbVtPBQAAxIk2DZSNGzdq4cKF+vGPfxy1/84779QLL7ygFStW6I033tBnn32m6667ri2nAgAA4kibBcqBAwdUVFSkJ554Ql27dnX2h8NhPfnkk5o5c6Yuu+wyXXDBBVq0aJHefvttrV+/vq2mAwAA4kibBUpxcbGGDh2q/Pz8qP3V1dVqaGiI2t+7d2/16NFDlZWVzZ6rvr5ekUgkagMAAO1XYlucdPny5dq8ebM2btzY5FgoFFJSUpLS0tKi9gcCAYVCoWbPV1ZWpqlTp7bFVAEAgIVcv4JSU1OjP/7xj3r66aeVnJzsyjlLS0sVDoedraamxpXzAgAAO7keKNXV1dq7d6/OP/98JSYmKjExUW+88YZmz56txMREBQIBHT58WHV1dVHPq62tVTAYbPacXq9XPp8vagMAAO2X6//Ec/nll+u9996L2jdq1Cj17t1bEydOVHZ2tjp27KiKigoVFhZKknbs2KFPPvlEeXl5bk8HAADEIdcDpUuXLjrvvPOi9nXq1EkZGRnO/tGjR6ukpETp6eny+Xz6/e9/r7y8PF100UVuTwcAAMShNrlJ9mQee+wxJSQkqLCwUPX19SooKNC8efNiMRUAAGAhjzHGxHoSrRWJROT3+xUOh7kf5X96Tlod6ykAiDO7pw+N9RRwhmnNz29+Fw8AALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsE5irCdgo56TVsd6CgDQ5uLx77rd04fGego4TbiCAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOsQKAAAwDoECgAAsA6BAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALCO64FSVlamCy+8UF26dFFmZqaGDx+uHTt2RI05dOiQiouLlZGRoc6dO6uwsFC1tbVuTwUAAMQp1wPljTfeUHFxsdavX681a9aooaFBV1xxhQ4ePOiMufPOO/XCCy9oxYoVeuONN/TZZ5/puuuuc3sqAAAgTiW6fcLy8vKox4sXL1ZmZqaqq6v105/+VOFwWE8++aSWLVumyy67TJK0aNEi9enTR+vXr9dFF13k9pQAAECcafN7UMLhsCQpPT1dklRdXa2Ghgbl5+c7Y3r37q0ePXqosrKy2XPU19crEolEbQAAoP1q00BpbGzU+PHjdfHFF+u8886TJIVCISUlJSktLS1qbCAQUCgUavY8ZWVl8vv9zpadnd2W0wYAADHWpoFSXFys999/X8uXLz+l85SWliocDjtbTU2NSzMEAAA2cv0elGPGjRunF198UevWrdP3vvc9Z38wGNThw4dVV1cXdRWltrZWwWCw2XN5vV55vd62mioAALCM61dQjDEaN26cnnvuOa1du1Y5OTlRxy+44AJ17NhRFRUVzr4dO3bok08+UV5entvTAQAAccj1KyjFxcVatmyZnn/+eXXp0sW5r8Tv9yslJUV+v1+jR49WSUmJ0tPT5fP59Pvf/155eXm8gwcAAEhqg0CZP3++JOmSSy6J2r9o0SLdcsstkqTHHntMCQkJKiwsVH19vQoKCjRv3jy3pwIAAOKU64FijDnpmOTkZM2dO1dz5851+48HAADtAL+LBwAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYB0CBQAAWIdAAQAA1iFQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFAAAYJ3EWE8AAICW6jlpdayn0Gq7pw+N9RTiEldQAACAdQgUAABgHQIFAABYh0ABAADWIVAAAIB1CBQAAGAdAgUAAFiHQAEAANbhg9oAAGhD8fjhclLsP2COKygAAMA6BAoAALAOgQIAAKxDoAAAAOvENFDmzp2rnj17Kjk5Wbm5udqwYUMspwMAACwRs0D561//qpKSEt17773avHmz+vXrp4KCAu3duzdWUwIAAJaIWaDMnDlTY8aM0ahRo3TOOedowYIFSk1N1VNPPRWrKQEAAEvE5HNQDh8+rOrqapWWljr7EhISlJ+fr8rKyibj6+vrVV9f7zwOh8OSpEgk0ibza6z/sk3OCwBAvGiLn7HHzmmMOenYmATKv//9bx09elSBQCBqfyAQ0IcffthkfFlZmaZOndpkf3Z2dpvNEQCAM5l/Vtud+4svvpDf7//GMXHxSbKlpaUqKSlxHjc2Nmrfvn3KyMiQx+OJ4czaTiQSUXZ2tmpqauTz+WI9ndOKtbN21n7mYO1n1tqNMfriiy+UlZV10rExCZRu3bqpQ4cOqq2tjdpfW1urYDDYZLzX65XX643al5aW1pZTtIbP5ztjvnGPx9pZ+5mGtbP2M8HJrpwcE5ObZJOSknTBBReooqLC2dfY2KiKigrl5eXFYkoAAMAiMfsnnpKSEo0cOVIDBw7UoEGDNGvWLB08eFCjRo2K1ZQAAIAlYhYoN9xwgz7//HNNmTJFoVBI/fv3V3l5eZMbZ89UXq9X9957b5N/2joTsHbWfqZh7awdTXlMS97rAwAAcBrxu3gAAIB1CBQAAGAdAgUAAFiHQAEAANYhUAAAgHUIFBdNmzZNgwcPVmpq6gk/6faTTz7R0KFDlZqaqszMTN199906cuRI1JjXX39d559/vrxer370ox9p8eLFTc4zd+5c9ezZU8nJycrNzdWGDRuijh86dEjFxcXKyMhQ586dVVhY2OSTe1syl1OxefNm/exnP1NaWpoyMjI0duxYHThwIGqMx+Npsi1fvjxqzOn6eripJWu36XvBTR999JGGDRumbt26yefzaciQIXrttdeixrTX170la2+Pr/vrr7/e7Gvq8Xi0ceNGSdLu3bubPb5+/fqoc61YsUK9e/dWcnKy+vbtq5deeinquDFGU6ZMUffu3ZWSkqL8/Hzt3Lkzasy+fftUVFQkn8+ntLQ0jR49usn//k7n2iXp3Xff1U9+8hMlJycrOztbM2bMaHKueFt7mzNwzZQpU8zMmTNNSUmJ8fv9TY4fOXLEnHfeeSY/P99s2bLFvPTSS6Zbt26mtLTUGfPPf/7TpKammpKSErNt2zYzZ84c06FDB1NeXu6MWb58uUlKSjJPPfWU+eCDD8yYMWNMWlqaqa2tdcbcfvvtJjs721RUVJhNmzaZiy66yAwePLhVczkVn376qenatau5/fbbzYcffmg2bNhgBg8ebAoLC6PGSTKLFi0ye/bscbavvvrqtH893NSStdv0veC2Xr16mZ///OfmnXfeMR999JG54447TGpqqtmzZ48zpj2+7i1Ze3t93evr66Neyz179pjbbrvN5OTkmMbGRmOMMbt27TKSzKuvvho17vDhw855/v73v5sOHTqYGTNmmG3btpl77rnHdOzY0bz33nvOmOnTpxu/329Wrlxp3nnnHXPNNdeYnJycqO+fK6+80vTr18+sX7/evPnmm+ZHP/qRufHGG2O29nA4bAKBgCkqKjLvv/++eeaZZ0xKSopZuHBhXK+9rREobWDRokXNBspLL71kEhISTCgUcvbNnz/f+Hw+U19fb4wxZsKECebcc8+Net4NN9xgCgoKnMeDBg0yxcXFzuOjR4+arKwsU1ZWZowxpq6uznTs2NGsWLHCGbN9+3YjyVRWVrZ4Lqdi4cKFJjMz0xw9etTZ9+677xpJZufOnc4+Sea555474XlO19fDTS1Zu03fC276/PPPjSSzbt06Z18kEjGSzJo1a5x97fF1b8na2+vrfrzDhw+b73znO+a+++5z9h0LlC1btpzweddff70ZOnRo1L7c3Fzz29/+1hhjTGNjowkGg+bhhx92jtfV1Rmv12ueeeYZY4wx27ZtM5LMxo0bnTEvv/yy8Xg85tNPP3Vjed+oubXPmzfPdO3aNerv1okTJ5qzzz7bedwe1u42/onnNKqsrFTfvn2jPi23oKBAkUhEH3zwgTMmPz8/6nkFBQWqrKyUJB0+fFjV1dVRYxISEpSfn++Mqa6uVkNDQ9SY3r17q0ePHs6YlszlVNTX1yspKUkJCf//WywlJUWS9NZbb0WNLS4uVrdu3TRo0CA99dRTMl/77MDT9fVwU0vWbtP3gpsyMjJ09tlna+nSpTp48KCOHDmihQsXKjMzUxdccEHU2Pb2urdk7e31dT/eqlWr9J///KfZX11yzTXXKDMzU0OGDNGqVauijp1s7bt27VIoFIoa4/f7lZubG/V3W1pamgYOHOiMyc/PV0JCgqqqqlxb44k0t/bKykr99Kc/VVJSkrOvoKBAO3bs0P79+50x8b52txEop1EoFGryUf7HHodCoW8cE4lE9NVXX+nf//63jh492uyYr58jKSmpyX0wx4852VxOxWWXXaZQKKSHH35Yhw8f1v79+zVp0iRJ0p49e5xx9913n5599lmtWbNGhYWFuuOOOzRnzhzn+On6eripJWu36XvBTR6PR6+++qq2bNmiLl26KDk5WTNnzlR5ebm6du3qjGuPr3tL1t5eX/fjPfnkkyooKND3vvc9Z1/nzp316KOPasWKFVq9erWGDBmi4cOHR0XKidb+9XUd2/dNYzIzM6OOJyYmKj09PWZrP5XXPZ7W7jYC5SQmTZp0whugjm0ffvhhrKd52rT063HuuedqyZIlevTRR5WamqpgMKicnBwFAoGoKwuTJ0/WxRdfrAEDBmjixImaMGGCHn744Riu8MTcXns8aenajTEqLi5WZmam3nzzTW3YsEHDhw/X1VdfHRWm7fF1b+na48m3+fvvX//6l1555RWNHj06an+3bt1UUlKi3NxcXXjhhZo+fbp+/etfx/3r/nUnWju+nZj9ssB4cdddd+mWW275xjE/+MEPWnSuYDDY5E77Y3fVB4NB5z+Pv9O+trZWPp9PKSkp6tChgzp06NDsmK+f4/Dhw6qrq4v6f1DHjznZXJrTmq/HTTfdpJtuukm1tbXq1KmTPB6PZs6c+Y1fr9zcXN1///2qr6+X1+s9bV+PlnBz7TZ9L7i59rVr1+rFF1/U/v375fP5JEnz5s3TmjVrtGTJEudK0vHaw+vekrW319f96xYtWqSMjAxdc801Jz1/bm6u1qxZ4zw+0dq/vq5j+7p37x41pn///s6YvXv3Rp3jyJEj2rdvX8zWfqJ1fX1NNq3dGrG9BaZ9OtlNsl+/037hwoXG5/OZQ4cOGWP+e4PceeedF/W8G2+8sckNcuPGjXMeHz161Hz3u99tcoPc3/72N2fMhx9+2OxNst80F7c9+eSTJjU11ezfv/+EYx544AHTtWtX5/Hp+nq0tePXbtP3gptWrVplEhISzBdffBG1/6yzzjLTpk074fPaw+vekrW319f9mMbGRpOTk2PuuuuuFo2/7bbbzIABA5zH119/vfnFL34RNSYvL6/JjaKPPPKIczwcDjd7o+imTZucMa+88kqb3yj6TWs/dpPs19+xVFpa2uQm2Xhde1shUFz08ccfmy1btpipU6eazp07my1btpgtW7Y4f2Ede4vhFVdcYbZu3WrKy8vNd77znWbfYnj33Xeb7du3m7lz5zb7FkOv12sWL15stm3bZsaOHWvS0tKi3hlw++23mx49epi1a9eaTZs2mby8PJOXl+ccb8lcTtWcOXNMdXW12bFjh/nLX/5iUlJSzJ///Gfn+KpVq8wTTzxh3nvvPbNz504zb948k5qaaqZMmXLavx5uO9nabfpecNPnn39uMjIyzHXXXWe2bt1qduzYYf70pz+Zjh07mq1btxpj2u/r3pK1t9fX/ZhXX33VSDLbt29vcmzx4sVm2bJlZvv27Wb79u1m2rRpJiEhwTz11FPOmL///e8mMTHRPPLII2b79u3m3nvvbfattmlpaeb555837777rhk2bFizb7UdMGCAqaqqMm+99Zbp1atXm7/V9pvWXldXZwKBgLn55pvN+++/b5YvX25SU1ObvM04XtfeVggUF40cOdJIarK99tprzpjdu3ebq666yqSkpJhu3bqZu+66yzQ0NESd57XXXjP9+/c3SUlJ5gc/+IFZtGhRkz9rzpw5pkePHiYpKckMGjTIrF+/Pur4V199Ze644w7TtWtXk5qaaq699tqoz6Fo6VxOxc0332zS09NNUlKS+fGPf2yWLl0adfzll182/fv3N507dzadOnUy/fr1MwsWLIh6e64xp+/r4aaTrd0Yu74X3LRx40ZzxRVXmPT0dNOlSxdz0UUXmZdeesk53p5f95Ot3Zj2+7ob898rPSf6rJXFixebPn36mNTUVOPz+cygQYOi3gZ9zLPPPmvOOussk5SUZM4991yzevXqqOONjY1m8uTJJhAIGK/Xay6//HKzY8eOqDH/+c9/zI033mg6d+5sfD6fGTVqVJMrW277prUbY8w777xjhgwZYrxer/nud79rpk+f3mRMvK69rXiM+dp7+wAAACwQn28pAAAA7RqBAgAArEOgAAAA6xAoAADAOgQKAACwDoECAACsQ6AAAADrECgAAMA6BAoAALAOgQIAAKxDoAAAAOv8P3710M2yj9q9AAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "t_lower = np.percentile(svm_model.decision_function(val_data2_scaled), 10)\n",
        "t_lower"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XiqDqZDrP99F",
        "outputId": "ce0e3399-8cde-42cc-ad81-9d84c2b97ce8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-98181.28538784165"
            ]
          },
          "metadata": {},
          "execution_count": 415
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "t_upper = np.percentile(svm_model.decision_function(val_data2_scaled), 90)\n",
        "t_upper"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iUP5SLkgQPqw",
        "outputId": "ead2e65c-3cf7-492a-d567-24c03b284e3c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-78901.40968468237"
            ]
          },
          "metadata": {},
          "execution_count": 416
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "t_lower = np.mean(svm_model.decision_function(val_data2_scaled)) - np.std(svm_model.decision_function(val_data2_scaled))\n",
        "t_lower"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4BATqnEtPIH-",
        "outputId": "3c29541f-4400-40d0-cada-95bd8b261686"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-0.006553227269295785"
            ]
          },
          "metadata": {},
          "execution_count": 357
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "t_upper = np.mean(svm_model.decision_function(val_data2_scaled)) +  np.std(svm_model.decision_function(val_data2_scaled))\n",
        "t_upper"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_X2yx6m4PPFA",
        "outputId": "aab83abc-c5e7-493d-84da-045002eaa3a3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-858.3465186945889"
            ]
          },
          "metadata": {},
          "execution_count": 407
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.mean(svm_model.decision_function(val_data_scaled)) - (np.std(val_data_scaled)/np.sqrt(len(val_data_scaled)))\n",
        "threshold"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "80I1FffWkKdw",
        "outputId": "704c5b18-1e19-4719-bdb7-10905a0f781c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-27.96456688358639"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test = svm_model.decision_function(test_data_scaled)\n",
        "pred = []\n",
        "for t in test:\n",
        "  if t >= t_upper:\n",
        "    pred.append(1)\n",
        "  else:\n",
        "    pred.append(0)\n"
      ],
      "metadata": {
        "id": "qkuUTq1okzgw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plt.hist(test)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "9LouPZWtU0Qc",
        "outputId": "580b536d-5171-469b-b4c7-85628b87517c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LAqO0KvbEPIE",
        "outputId": "b543f3d8-605a-481a-d1f5-bd8089aca944"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0]"
            ]
          },
          "metadata": {},
          "execution_count": 419
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(gt, pred, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "F2OiBiRjlb4g",
        "outputId": "b4b62ae0-2dfe-4c9e-9a5e-22d28048f728"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.67      0.99      0.80        71\n",
            "      attack       0.94      0.31      0.46        49\n",
            "\n",
            "    accuracy                           0.71       120\n",
            "   macro avg       0.81      0.65      0.63       120\n",
            "weighted avg       0.78      0.71      0.66       120\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "##RBM"
      ],
      "metadata": {
        "id": "1YdkOoCDOZEu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "import numpy as np\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "\n",
        "# Define the RBM model\n",
        "class RBM(nn.Module):\n",
        "    def __init__(self, visible_units, hidden_units):\n",
        "        super(RBM, self).__init__()\n",
        "        self.W = nn.Parameter(torch.randn(hidden_units, visible_units))\n",
        "        self.v_bias = nn.Parameter(torch.randn(visible_units))\n",
        "        self.h_bias = nn.Parameter(torch.randn(hidden_units))\n",
        "\n",
        "    def forward(self, x):\n",
        "        hidden_prob = torch.sigmoid(torch.matmul(x, self.W.t()) + self.h_bias)\n",
        "        hidden_state = torch.bernoulli(hidden_prob)\n",
        "        visible_prob = torch.sigmoid(torch.matmul(hidden_state, self.W) + self.v_bias)\n",
        "        return visible_prob, hidden_state\n",
        "\n",
        "    def generate_samples(self, num_samples):\n",
        "        hidden_state = torch.bernoulli(torch.sigmoid(self.h_bias))\n",
        "        samples = []\n",
        "        for _ in range(num_samples):\n",
        "            visible_prob = torch.sigmoid(torch.matmul(hidden_state, self.W.t()) + self.v_bias)\n",
        "            visible_state = torch.bernoulli(visible_prob)\n",
        "            hidden_prob = torch.sigmoid(torch.matmul(visible_state, self.W) + self.h_bias)\n",
        "            hidden_state = torch.bernoulli(hidden_prob)\n",
        "            samples.append(visible_state)\n",
        "        return torch.stack(samples)\n",
        "\n",
        "# Set random seed for reproducibility\n",
        "torch.manual_seed(42)\n",
        "np.random.seed(42)\n",
        "# Convert the training data to Torch tensors\n",
        "train_data = torch.from_numpy(train_data_scaled).float()\n",
        "\n",
        "# Create an instance of the RBM model\n",
        "rbm_model = RBM(visible_units=50, hidden_units=128)  # Adjust the number of units as needed\n",
        "\n",
        "# Set the optimization criterion and optimizer\n",
        "criterion = nn.MSELoss()\n",
        "optimizer = optim.SGD(rbm_model.parameters(), lr=0.01)\n",
        "\n",
        "# Training loop\n",
        "num_epochs = 100\n",
        "batch_size = 64\n",
        "for epoch in range(num_epochs):\n",
        "    # Shuffle the training data\n",
        "    indices = torch.randperm(train_data.size(0))\n",
        "    shuffled_data = train_data[indices]\n",
        "\n",
        "    # Mini-batch training\n",
        "    for i in range(0, train_data.size(0), batch_size):\n",
        "        inputs = shuffled_data[i:i+batch_size]\n",
        "\n",
        "        # Forward pass\n",
        "        visible_prob, _ = rbm_model(inputs)\n",
        "\n",
        "        # Contrastive Divergence (CD) algorithm\n",
        "        hidden_prob = torch.sigmoid(torch.matmul(visible_prob, rbm_model.W.t()) + rbm_model.h_bias)\n",
        "        hidden_state = torch.bernoulli(hidden_prob)\n",
        "        visible_recon_prob, _ = rbm_model(torch.matmul(hidden_state, rbm_model.W) + rbm_model.v_bias)\n",
        "\n",
        "        # Update weights and biases\n",
        "        optimizer.zero_grad()\n",
        "        loss = criterion(inputs, visible_recon_prob)\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "\n",
        "    # Print the loss for monitoring\n",
        "    print(f\"Epoch [{epoch+1}/{num_epochs}], Loss: {loss.item():.4f}\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "slEk-RUhOaDa",
        "outputId": "d218f48b-609f-4105-b40e-8e4e01399fa2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch [1/100], Loss: 0.4093\n",
            "Epoch [2/100], Loss: 0.4017\n",
            "Epoch [3/100], Loss: 0.3999\n",
            "Epoch [4/100], Loss: 0.4004\n",
            "Epoch [5/100], Loss: 0.3977\n",
            "Epoch [6/100], Loss: 0.3935\n",
            "Epoch [7/100], Loss: 0.3927\n",
            "Epoch [8/100], Loss: 0.3853\n",
            "Epoch [9/100], Loss: 0.3911\n",
            "Epoch [10/100], Loss: 0.3928\n",
            "Epoch [11/100], Loss: 0.3867\n",
            "Epoch [12/100], Loss: 0.3893\n",
            "Epoch [13/100], Loss: 0.3761\n",
            "Epoch [14/100], Loss: 0.3680\n",
            "Epoch [15/100], Loss: 0.3699\n",
            "Epoch [16/100], Loss: 0.3766\n",
            "Epoch [17/100], Loss: 0.3661\n",
            "Epoch [18/100], Loss: 0.3726\n",
            "Epoch [19/100], Loss: 0.3545\n",
            "Epoch [20/100], Loss: 0.3678\n",
            "Epoch [21/100], Loss: 0.3690\n",
            "Epoch [22/100], Loss: 0.3554\n",
            "Epoch [23/100], Loss: 0.3530\n",
            "Epoch [24/100], Loss: 0.3492\n",
            "Epoch [25/100], Loss: 0.3522\n",
            "Epoch [26/100], Loss: 0.3456\n",
            "Epoch [27/100], Loss: 0.3409\n",
            "Epoch [28/100], Loss: 0.3375\n",
            "Epoch [29/100], Loss: 0.3378\n",
            "Epoch [30/100], Loss: 0.3377\n",
            "Epoch [31/100], Loss: 0.3294\n",
            "Epoch [32/100], Loss: 0.3409\n",
            "Epoch [33/100], Loss: 0.3205\n",
            "Epoch [34/100], Loss: 0.3284\n",
            "Epoch [35/100], Loss: 0.3198\n",
            "Epoch [36/100], Loss: 0.3166\n",
            "Epoch [37/100], Loss: 0.3236\n",
            "Epoch [38/100], Loss: 0.3098\n",
            "Epoch [39/100], Loss: 0.3030\n",
            "Epoch [40/100], Loss: 0.3031\n",
            "Epoch [41/100], Loss: 0.3019\n",
            "Epoch [42/100], Loss: 0.3014\n",
            "Epoch [43/100], Loss: 0.2990\n",
            "Epoch [44/100], Loss: 0.2959\n",
            "Epoch [45/100], Loss: 0.2890\n",
            "Epoch [46/100], Loss: 0.2927\n",
            "Epoch [47/100], Loss: 0.2903\n",
            "Epoch [48/100], Loss: 0.2881\n",
            "Epoch [49/100], Loss: 0.2796\n",
            "Epoch [50/100], Loss: 0.2808\n",
            "Epoch [51/100], Loss: 0.2721\n",
            "Epoch [52/100], Loss: 0.2763\n",
            "Epoch [53/100], Loss: 0.2702\n",
            "Epoch [54/100], Loss: 0.2740\n",
            "Epoch [55/100], Loss: 0.2695\n",
            "Epoch [56/100], Loss: 0.2696\n",
            "Epoch [57/100], Loss: 0.2654\n",
            "Epoch [58/100], Loss: 0.2706\n",
            "Epoch [59/100], Loss: 0.2623\n",
            "Epoch [60/100], Loss: 0.2580\n",
            "Epoch [61/100], Loss: 0.2604\n",
            "Epoch [62/100], Loss: 0.2571\n",
            "Epoch [63/100], Loss: 0.2644\n",
            "Epoch [64/100], Loss: 0.2606\n",
            "Epoch [65/100], Loss: 0.2549\n",
            "Epoch [66/100], Loss: 0.2514\n",
            "Epoch [67/100], Loss: 0.2467\n",
            "Epoch [68/100], Loss: 0.2561\n",
            "Epoch [69/100], Loss: 0.2505\n",
            "Epoch [70/100], Loss: 0.2534\n",
            "Epoch [71/100], Loss: 0.2457\n",
            "Epoch [72/100], Loss: 0.2516\n",
            "Epoch [73/100], Loss: 0.2481\n",
            "Epoch [74/100], Loss: 0.2427\n",
            "Epoch [75/100], Loss: 0.2462\n",
            "Epoch [76/100], Loss: 0.2494\n",
            "Epoch [77/100], Loss: 0.2466\n",
            "Epoch [78/100], Loss: 0.2516\n",
            "Epoch [79/100], Loss: 0.2422\n",
            "Epoch [80/100], Loss: 0.2456\n",
            "Epoch [81/100], Loss: 0.2438\n",
            "Epoch [82/100], Loss: 0.2517\n",
            "Epoch [83/100], Loss: 0.2430\n",
            "Epoch [84/100], Loss: 0.2389\n",
            "Epoch [85/100], Loss: 0.2382\n",
            "Epoch [86/100], Loss: 0.2412\n",
            "Epoch [87/100], Loss: 0.2461\n",
            "Epoch [88/100], Loss: 0.2308\n",
            "Epoch [89/100], Loss: 0.2426\n",
            "Epoch [90/100], Loss: 0.2411\n",
            "Epoch [91/100], Loss: 0.2336\n",
            "Epoch [92/100], Loss: 0.2395\n",
            "Epoch [93/100], Loss: 0.2397\n",
            "Epoch [94/100], Loss: 0.2328\n",
            "Epoch [95/100], Loss: 0.2342\n",
            "Epoch [96/100], Loss: 0.2393\n",
            "Epoch [97/100], Loss: 0.2397\n",
            "Epoch [98/100], Loss: 0.2386\n",
            "Epoch [99/100], Loss: 0.2318\n",
            "Epoch [100/100], Loss: 0.2342\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Convert the test data to Torch tensors\n",
        "val_data = torch.from_numpy(val_data1_scaled).float()\n",
        "\n",
        "# Set the RBM model to evaluation mode\n",
        "rbm_model.eval()\n",
        "\n",
        "# Calculate the reconstruction errors for the test data\n",
        "reconstruction_errors = []\n",
        "for i in range(val_data.size(0)):\n",
        "    inputs = val_data[i:i+1]\n",
        "    visible_recon_prob, _ = rbm_model(inputs)\n",
        "    error = criterion(inputs, visible_recon_prob)\n",
        "    reconstruction_errors.append(error.item())\n",
        "\n"
      ],
      "metadata": {
        "id": "LCkIZrnWGbHk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Define a threshold for classifying anomalies\n",
        "reconstruction_error_threshold = np.mean(reconstruction_errors) #+ 2*np.std(reconstruction_errors)\n",
        "reconstruction_error_threshold"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DfDo4LAyauBd",
        "outputId": "e87e744f-5930-432b-acf9-a266bfaac53e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.4100679231186708"
            ]
          },
          "metadata": {},
          "execution_count": 200
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "\n",
        "# Convert the test data to Torch tensors\n",
        "test_data = torch.from_numpy(test_data_scaled).float()\n",
        "\n",
        "# Set the RBM model to evaluation mode\n",
        "rbm_model.eval()\n",
        "\n",
        "# Calculate the reconstruction errors for the test data\n",
        "reconstruction_errors = []\n",
        "for i in range(test_data.size(0)):\n",
        "    inputs = test_data[i:i+1]\n",
        "    visible_recon_prob, _ = rbm_model(inputs)\n",
        "    error = criterion(inputs, visible_recon_prob)\n",
        "    reconstruction_errors.append(error.item())\n",
        "\n",
        "# Classify instances as normal (0) or anomaly (1) based on the threshold\n",
        "predictions = [0 if error <= reconstruction_error_threshold else 1 for error in reconstruction_errors]\n",
        "\n",
        "# Print the predictions\n",
        "print(\"Predictions:\")\n",
        "print(predictions)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4b1ZazQ9Dbsu",
        "outputId": "d58b73f8-d014-4fe9-b36b-222944a49773"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Predictions:\n",
            "[0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 1, 1, 0, 0, 0, 1, 1, 0, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 0]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "predictions[]"
      ],
      "metadata": {
        "id": "LqCP5bSCP6kT"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "np.mean(reconstruction_errors)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xea9H6kuws0F",
        "outputId": "df3e812e-d2d5-44b7-9a31-1f33ae43bea7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.41441892882188164"
            ]
          },
          "metadata": {},
          "execution_count": 96
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "reconstruction_error_threshold"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sMRSH349GXHQ",
        "outputId": "f917c1c3-fba6-4f20-e5c5-6d5d88376d79"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.4046303614632373"
            ]
          },
          "metadata": {},
          "execution_count": 86
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "gt = predictions['Gt']"
      ],
      "metadata": {
        "id": "mVtbHWuzGEhB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(gt, predictions, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Fy1XvdBfFxec",
        "outputId": "0896bb23-7fb1-40e2-c548-aeb5aa0c5a9c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.55      0.49      0.52        71\n",
            "      attack       0.36      0.41      0.38        49\n",
            "\n",
            "    accuracy                           0.46       120\n",
            "   macro avg       0.45      0.45      0.45       120\n",
            "weighted avg       0.47      0.46      0.46       120\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Overcomlete AE 2nd version"
      ],
      "metadata": {
        "id": "4MtgiqS_p8v_"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras.models import Model\n",
        "from tensorflow.keras.layers import Input, Dense, Dropout, BatchNormalization, concatenate\n",
        "\n",
        "def OAE(input_dim, hidden_dim, output_dim):\n",
        "    inputs = Input(shape=(input_dim,))\n",
        "\n",
        "    # Encoder\n",
        "    encoder1 = Dense(hidden_dim, activation='relu')(inputs)\n",
        "    encoder1 = Dropout(0.2)(encoder1)\n",
        "    encoder1 = BatchNormalization()(encoder1)\n",
        "\n",
        "    encoder2 = Dense(64, activation='relu')(encoder1)\n",
        "    encoder2 = Dropout(0.2)(encoder2)\n",
        "    encoder2 = BatchNormalization()(encoder2)\n",
        "\n",
        "    encoder3 = Dense(32, activation='relu')(encoder2)\n",
        "    encoder3 = Dropout(0.2)(encoder3)\n",
        "    encoder3 = BatchNormalization()(encoder3)\n",
        "\n",
        "    decoder2 = Dense(64, activation='relu')(encoder3)\n",
        "    decoder2 = Dropout(0.2)(decoder2)\n",
        "    decoder2 = BatchNormalization()(decoder2)\n",
        "\n",
        "    decoder1 = Dense(hidden_dim, activation='relu')(decoder2)\n",
        "    decoder1 = Dropout(0.2)(decoder1)\n",
        "    decoder1 = BatchNormalization()(decoder1)\n",
        "\n",
        "    # Output\n",
        "    output = Dense(output_dim, activation='sigmoid')(decoder1)\n",
        "\n",
        "    model = Model(inputs=inputs, outputs=output)\n",
        "    return model\n",
        "\n",
        "# Define your dataset and data preprocessing\n",
        "# ...\n",
        "\n",
        "# Define hyperparameters\n",
        "input_dim = train_data.shape[1]\n",
        "hidden_dim = 128\n",
        "output_dim = train_data.shape[1]\n",
        "learning_rate = 0.001\n",
        "batch_size = 64\n",
        "num_epochs = 100\n",
        "\n",
        "# Instantiate the UNet model\n",
        "model = OAE(input_dim, hidden_dim, output_dim)\n",
        "\n",
        "# Define loss function and optimizer\n",
        "loss_fn = tf.keras.losses.MeanSquaredError()\n",
        "optimizer = tf.keras.optimizers.Adam(learning_rate=learning_rate)\n",
        "\n",
        "# Compile the model\n",
        "model.compile(optimizer=optimizer, loss=loss_fn)\n",
        "\n",
        "# Train the model\n",
        "model.fit(train_data_scaled, train_data_scaled, epochs=num_epochs, batch_size=batch_size, shuffle=True, validation_data=(val_data1_scaled, val_data1_scaled))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Mzrv8ALpqAEy",
        "outputId": "c2db77b4-3e8b-497b-f748-6932262e4caa"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "71/71 [==============================] - 6s 11ms/step - loss: 0.1906 - val_loss: 0.1765\n",
            "Epoch 2/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0636 - val_loss: 0.2904\n",
            "Epoch 3/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0124 - val_loss: 0.3305\n",
            "Epoch 4/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0072 - val_loss: 0.3395\n",
            "Epoch 5/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0056 - val_loss: 0.3516\n",
            "Epoch 6/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0048 - val_loss: 0.3628\n",
            "Epoch 7/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0044 - val_loss: 0.3829\n",
            "Epoch 8/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0040 - val_loss: 0.3987\n",
            "Epoch 9/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0036 - val_loss: 0.4034\n",
            "Epoch 10/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0033 - val_loss: 0.4096\n",
            "Epoch 11/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 0.0030 - val_loss: 0.4133\n",
            "Epoch 12/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0029 - val_loss: 0.4038\n",
            "Epoch 13/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0028 - val_loss: 0.3998\n",
            "Epoch 14/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0026 - val_loss: 0.3936\n",
            "Epoch 15/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0025 - val_loss: 0.3849\n",
            "Epoch 16/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0024 - val_loss: 0.3872\n",
            "Epoch 17/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0023 - val_loss: 0.3844\n",
            "Epoch 18/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0024 - val_loss: 0.3829\n",
            "Epoch 19/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0022 - val_loss: 0.3842\n",
            "Epoch 20/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0022 - val_loss: 0.3807\n",
            "Epoch 21/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0021 - val_loss: 0.3765\n",
            "Epoch 22/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0021 - val_loss: 0.3742\n",
            "Epoch 23/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0021 - val_loss: 0.3760\n",
            "Epoch 24/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0020 - val_loss: 0.3720\n",
            "Epoch 25/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0020 - val_loss: 0.3758\n",
            "Epoch 26/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0020 - val_loss: 0.3747\n",
            "Epoch 27/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0019 - val_loss: 0.3697\n",
            "Epoch 28/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0019 - val_loss: 0.3556\n",
            "Epoch 29/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0019 - val_loss: 0.3519\n",
            "Epoch 30/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0019 - val_loss: 0.3525\n",
            "Epoch 31/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0018 - val_loss: 0.3390\n",
            "Epoch 32/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0018 - val_loss: 0.3346\n",
            "Epoch 33/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0018 - val_loss: 0.2743\n",
            "Epoch 34/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0018 - val_loss: 0.2328\n",
            "Epoch 35/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0018 - val_loss: 0.2440\n",
            "Epoch 36/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0018 - val_loss: 0.2484\n",
            "Epoch 37/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0017 - val_loss: 0.2272\n",
            "Epoch 38/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 0.0016 - val_loss: 0.2304\n",
            "Epoch 39/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0017 - val_loss: 0.2467\n",
            "Epoch 40/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0017 - val_loss: 0.2710\n",
            "Epoch 41/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0016 - val_loss: 0.2213\n",
            "Epoch 42/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0016 - val_loss: 0.2711\n",
            "Epoch 43/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 0.0016 - val_loss: 0.2724\n",
            "Epoch 44/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0017 - val_loss: 0.2603\n",
            "Epoch 45/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0016 - val_loss: 0.2807\n",
            "Epoch 46/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0017 - val_loss: 0.2983\n",
            "Epoch 47/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0015 - val_loss: 0.2924\n",
            "Epoch 48/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0015 - val_loss: 0.2551\n",
            "Epoch 49/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0016 - val_loss: 0.2879\n",
            "Epoch 50/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0016 - val_loss: 0.3023\n",
            "Epoch 51/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0016 - val_loss: 0.2993\n",
            "Epoch 52/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0015 - val_loss: 0.2900\n",
            "Epoch 53/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0015 - val_loss: 0.2418\n",
            "Epoch 54/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0015 - val_loss: 0.2492\n",
            "Epoch 55/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0015 - val_loss: 0.2627\n",
            "Epoch 56/100\n",
            "71/71 [==============================] - 0s 6ms/step - loss: 0.0015 - val_loss: 0.3050\n",
            "Epoch 57/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0016 - val_loss: 0.2589\n",
            "Epoch 58/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0015 - val_loss: 0.2835\n",
            "Epoch 59/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 0.0014 - val_loss: 0.2768\n",
            "Epoch 60/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 0.0015 - val_loss: 0.3000\n",
            "Epoch 61/100\n",
            "71/71 [==============================] - 1s 16ms/step - loss: 0.0015 - val_loss: 0.2586\n",
            "Epoch 62/100\n",
            "71/71 [==============================] - 1s 16ms/step - loss: 0.0014 - val_loss: 0.2797\n",
            "Epoch 63/100\n",
            "71/71 [==============================] - 1s 16ms/step - loss: 0.0015 - val_loss: 0.2354\n",
            "Epoch 64/100\n",
            "71/71 [==============================] - 1s 17ms/step - loss: 0.0014 - val_loss: 0.2371\n",
            "Epoch 65/100\n",
            "71/71 [==============================] - 1s 15ms/step - loss: 0.0014 - val_loss: 0.2422\n",
            "Epoch 66/100\n",
            "71/71 [==============================] - 1s 9ms/step - loss: 0.0014 - val_loss: 0.2438\n",
            "Epoch 67/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 0.0014 - val_loss: 0.2327\n",
            "Epoch 68/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 0.0014 - val_loss: 0.2375\n",
            "Epoch 69/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0014 - val_loss: 0.2575\n",
            "Epoch 70/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 0.0014 - val_loss: 0.2529\n",
            "Epoch 71/100\n",
            "71/71 [==============================] - 1s 13ms/step - loss: 0.0014 - val_loss: 0.2799\n",
            "Epoch 72/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 0.0014 - val_loss: 0.2748\n",
            "Epoch 73/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0013 - val_loss: 0.2758\n",
            "Epoch 74/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0014 - val_loss: 0.2740\n",
            "Epoch 75/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0013 - val_loss: 0.2665\n",
            "Epoch 76/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0013 - val_loss: 0.2632\n",
            "Epoch 77/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0012 - val_loss: 0.2536\n",
            "Epoch 78/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0013 - val_loss: 0.2769\n",
            "Epoch 79/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0013 - val_loss: 0.2482\n",
            "Epoch 80/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0013 - val_loss: 0.2478\n",
            "Epoch 81/100\n",
            "71/71 [==============================] - 1s 12ms/step - loss: 0.0013 - val_loss: 0.2676\n",
            "Epoch 82/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0012 - val_loss: 0.2472\n",
            "Epoch 83/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0013 - val_loss: 0.2442\n",
            "Epoch 84/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 0.0013 - val_loss: 0.2405\n",
            "Epoch 85/100\n",
            "71/71 [==============================] - 1s 11ms/step - loss: 0.0013 - val_loss: 0.2518\n",
            "Epoch 86/100\n",
            "71/71 [==============================] - 1s 10ms/step - loss: 0.0012 - val_loss: 0.2609\n",
            "Epoch 87/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0013 - val_loss: 0.2831\n",
            "Epoch 88/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0013 - val_loss: 0.2753\n",
            "Epoch 89/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0013 - val_loss: 0.2716\n",
            "Epoch 90/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0013 - val_loss: 0.2549\n",
            "Epoch 91/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0013 - val_loss: 0.2078\n",
            "Epoch 92/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0012 - val_loss: 0.1996\n",
            "Epoch 93/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0012 - val_loss: 0.1986\n",
            "Epoch 94/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0012 - val_loss: 0.1954\n",
            "Epoch 95/100\n",
            "71/71 [==============================] - 1s 8ms/step - loss: 0.0013 - val_loss: 0.2095\n",
            "Epoch 96/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0012 - val_loss: 0.2005\n",
            "Epoch 97/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0012 - val_loss: 0.2073\n",
            "Epoch 98/100\n",
            "71/71 [==============================] - 1s 7ms/step - loss: 0.0012 - val_loss: 0.2071\n",
            "Epoch 99/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0012 - val_loss: 0.1715\n",
            "Epoch 100/100\n",
            "71/71 [==============================] - 0s 7ms/step - loss: 0.0011 - val_loss: 0.1696\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7ff23a509450>"
            ]
          },
          "metadata": {},
          "execution_count": 123
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Anomaly detection\n",
        "anomaly_scores = []\n",
        "for inputs in val_data2_scaled:\n",
        "    inputs = tf.expand_dims(inputs, 0)  # Add batch dimension\n",
        "    outputs = model.predict(inputs)\n",
        "    loss = loss_fn(outputs, inputs)\n",
        "    anomaly_scores.append(loss.numpy())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_UiBgwhYqdPb",
        "outputId": "2a7a72b8-e451-4e1b-c04d-e8e36e6d604b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 62ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 66ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 115ms/step\n",
            "1/1 [==============================] - 0s 66ms/step\n",
            "1/1 [==============================] - 0s 72ms/step\n",
            "1/1 [==============================] - 0s 89ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 89ms/step\n",
            "1/1 [==============================] - 0s 91ms/step\n",
            "1/1 [==============================] - 0s 80ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 98ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 70ms/step\n",
            "1/1 [==============================] - 0s 106ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 58ms/step\n",
            "1/1 [==============================] - 0s 93ms/step\n",
            "1/1 [==============================] - 0s 59ms/step\n",
            "1/1 [==============================] - 0s 59ms/step\n",
            "1/1 [==============================] - 0s 133ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 109ms/step\n",
            "1/1 [==============================] - 0s 83ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 48ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 63ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 73ms/step\n",
            "1/1 [==============================] - 0s 69ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 67ms/step\n",
            "1/1 [==============================] - 0s 56ms/step\n",
            "1/1 [==============================] - 0s 59ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 87ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 83ms/step\n",
            "1/1 [==============================] - 0s 114ms/step\n",
            "1/1 [==============================] - 0s 81ms/step\n",
            "1/1 [==============================] - 0s 130ms/step\n",
            "1/1 [==============================] - 0s 116ms/step\n",
            "1/1 [==============================] - 0s 91ms/step\n",
            "1/1 [==============================] - 0s 118ms/step\n",
            "1/1 [==============================] - 0s 161ms/step\n",
            "1/1 [==============================] - 0s 126ms/step\n",
            "1/1 [==============================] - 0s 59ms/step\n",
            "1/1 [==============================] - 0s 104ms/step\n",
            "1/1 [==============================] - 0s 124ms/step\n",
            "1/1 [==============================] - 0s 309ms/step\n",
            "1/1 [==============================] - 0s 261ms/step\n",
            "1/1 [==============================] - 0s 128ms/step\n",
            "1/1 [==============================] - 0s 132ms/step\n",
            "1/1 [==============================] - 0s 158ms/step\n",
            "1/1 [==============================] - 0s 69ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 50ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 48ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 89ms/step\n",
            "1/1 [==============================] - 0s 105ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 105ms/step\n",
            "1/1 [==============================] - 0s 91ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 64ms/step\n",
            "1/1 [==============================] - 0s 54ms/step\n",
            "1/1 [==============================] - 0s 88ms/step\n",
            "1/1 [==============================] - 0s 101ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 50ms/step\n",
            "1/1 [==============================] - 0s 95ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 73ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 54ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 108ms/step\n",
            "1/1 [==============================] - 0s 177ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 54ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 48ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 78ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 56ms/step\n",
            "1/1 [==============================] - 0s 58ms/step\n",
            "1/1 [==============================] - 0s 66ms/step\n",
            "1/1 [==============================] - 0s 55ms/step\n",
            "1/1 [==============================] - 0s 58ms/step\n",
            "1/1 [==============================] - 0s 59ms/step\n",
            "1/1 [==============================] - 0s 60ms/step\n",
            "1/1 [==============================] - 0s 57ms/step\n",
            "1/1 [==============================] - 0s 62ms/step\n",
            "1/1 [==============================] - 0s 60ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 64ms/step\n",
            "1/1 [==============================] - 0s 57ms/step\n",
            "1/1 [==============================] - 0s 71ms/step\n",
            "1/1 [==============================] - 0s 59ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 68ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 57ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n",
            "1/1 [==============================] - 0s 65ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 58ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 50ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 50ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 48ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 50ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 66ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 56ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 35ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 33ms/step\n",
            "1/1 [==============================] - 0s 34ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 32ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "np.percentile(anomaly_scores, 90)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wRNP_NeeWh39",
        "outputId": "087f61e0-1060-44bf-e020-e21758ca1a51"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.2639789925966438"
            ]
          },
          "metadata": {},
          "execution_count": 126
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.mean(anomaly_scores) + (np.std(anomaly_scores)/np.sqrt(len(anomaly_scores)))\n",
        "threshold"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Res6xvBgXHQ_",
        "outputId": "3641de58-4a9e-4475-ef1a-2d02bd38a00d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.1671721901572071"
            ]
          },
          "metadata": {},
          "execution_count": 134
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.mean(anomaly_scores) + (np.std(anomaly_scores)/len(anomaly_scores)) # Adjust threshold based on your data and requirements\n",
        "threshold"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oEop35KNqfC4",
        "outputId": "e45555c7-5c8b-4830-dfa4-40aefb89f8d8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.16411094246416494"
            ]
          },
          "metadata": {},
          "execution_count": 127
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Anomaly detection\n",
        "anomaly_scores_test = []\n",
        "for inputs in test_data_scaled:\n",
        "    inputs = tf.expand_dims(inputs, 0)  # Add batch dimension\n",
        "    outputs = model.predict(inputs)\n",
        "    loss = loss_fn(outputs, inputs)\n",
        "    anomaly_scores_test.append(loss.numpy())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uzW62QPXqi3e",
        "outputId": "11021684-7f30-4c19-d975-88e0f9218626"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 104ms/step\n",
            "1/1 [==============================] - 0s 106ms/step\n",
            "1/1 [==============================] - 0s 152ms/step\n",
            "1/1 [==============================] - 0s 109ms/step\n",
            "1/1 [==============================] - 0s 109ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 94ms/step\n",
            "1/1 [==============================] - 0s 116ms/step\n",
            "1/1 [==============================] - 0s 144ms/step\n",
            "1/1 [==============================] - 0s 111ms/step\n",
            "1/1 [==============================] - 0s 125ms/step\n",
            "1/1 [==============================] - 0s 154ms/step\n",
            "1/1 [==============================] - 0s 73ms/step\n",
            "1/1 [==============================] - 0s 65ms/step\n",
            "1/1 [==============================] - 0s 75ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n",
            "1/1 [==============================] - 0s 150ms/step\n",
            "1/1 [==============================] - 0s 149ms/step\n",
            "1/1 [==============================] - 0s 95ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 72ms/step\n",
            "1/1 [==============================] - 0s 45ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 124ms/step\n",
            "1/1 [==============================] - 0s 54ms/step\n",
            "1/1 [==============================] - 0s 119ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 73ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 54ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 57ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 84ms/step\n",
            "1/1 [==============================] - 0s 52ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 166ms/step\n",
            "1/1 [==============================] - 0s 140ms/step\n",
            "1/1 [==============================] - 0s 142ms/step\n",
            "1/1 [==============================] - 0s 137ms/step\n",
            "1/1 [==============================] - 0s 61ms/step\n",
            "1/1 [==============================] - 0s 110ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 46ms/step\n",
            "1/1 [==============================] - 0s 55ms/step\n",
            "1/1 [==============================] - 0s 55ms/step\n",
            "1/1 [==============================] - 0s 84ms/step\n",
            "1/1 [==============================] - 0s 93ms/step\n",
            "1/1 [==============================] - 0s 72ms/step\n",
            "1/1 [==============================] - 0s 75ms/step\n",
            "1/1 [==============================] - 0s 53ms/step\n",
            "1/1 [==============================] - 0s 37ms/step\n",
            "1/1 [==============================] - 0s 92ms/step\n",
            "1/1 [==============================] - 0s 97ms/step\n",
            "1/1 [==============================] - 0s 133ms/step\n",
            "1/1 [==============================] - 0s 86ms/step\n",
            "1/1 [==============================] - 0s 96ms/step\n",
            "1/1 [==============================] - 0s 111ms/step\n",
            "1/1 [==============================] - 0s 96ms/step\n",
            "1/1 [==============================] - 0s 114ms/step\n",
            "1/1 [==============================] - 0s 128ms/step\n",
            "1/1 [==============================] - 0s 64ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 39ms/step\n",
            "1/1 [==============================] - 0s 44ms/step\n",
            "1/1 [==============================] - 0s 47ms/step\n",
            "1/1 [==============================] - 0s 78ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 38ms/step\n",
            "1/1 [==============================] - 0s 40ms/step\n",
            "1/1 [==============================] - 0s 42ms/step\n",
            "1/1 [==============================] - 0s 43ms/step\n",
            "1/1 [==============================] - 0s 48ms/step\n",
            "1/1 [==============================] - 0s 49ms/step\n",
            "1/1 [==============================] - 0s 41ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 29ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 31ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 36ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Thresholding and labeling anomalies\n",
        "anomalies = [1 if score > threshold else 0 for score in anomaly_scores_test]"
      ],
      "metadata": {
        "id": "znfXtGJ3qogl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "anomalies"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 165
        },
        "id": "Y6T4YFxMr9ow",
        "outputId": "78f02606-f5b2-40db-89ae-ca45e671ceba"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "NameError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-1-8f02c20d29a7>\u001b[0m in \u001b[0;36m<cell line: 1>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0manomalies\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;31mNameError\u001b[0m: name 'anomalies' is not defined"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(gt, anomalies, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "id": "9G02gFhu1lD1"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## LSTM"
      ],
      "metadata": {
        "id": "vvNbi2l8lX64"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "test = pd.read_csv('drive/MyDrive/Tesi/Experiment/Data/Test/Refined/test_final.csv', sep=',')\n",
        "test.drop(labels='Unnamed: 0', axis=1, inplace=True)\n",
        "test.drop(labels='Time', axis=1, inplace=True)"
      ],
      "metadata": {
        "id": "UJ3omDfFqyxo"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Step 1: Get the last 10 rows\n",
        "last_10_rows = test.tail(10)\n",
        "\n",
        "# Step 2: Remove the last 10 rows from the original DataFrame\n",
        "test_lstm = test[:-10]\n",
        "\n",
        "# Step 3: Insert the last 10 rows at the top\n",
        "test_lstm = pd.concat([last_10_rows, test_lstm], ignore_index=True)"
      ],
      "metadata": {
        "id": "A_4bSVEcquAq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test_lstm"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 522
        },
        "id": "dr4x2C2rrB3X",
        "outputId": "acae0193-9040-41da-cc62-bb95159f6937"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     idle_node1_delta  iowait_node1_delta  user_node1_delta  \\\n",
              "0              441.64                0.44             23.35   \n",
              "1              442.77                0.43             22.99   \n",
              "2              440.40                0.47             25.34   \n",
              "3              441.59                0.39             23.98   \n",
              "4              444.86                0.41             21.26   \n",
              "..                ...                 ...               ...   \n",
              "115            443.05                0.43             22.67   \n",
              "116            441.79                0.44             23.15   \n",
              "117            442.36                0.44             22.97   \n",
              "118            443.61                0.43             21.65   \n",
              "119            442.37                0.47             23.06   \n",
              "\n",
              "     system_node1_delta  irq_node1_delta  softirq_node1_delta  \\\n",
              "0                  8.16              0.0                 0.87   \n",
              "1                  7.19              0.0                 0.86   \n",
              "2                  7.46              0.0                 1.02   \n",
              "3                  7.60              0.0                 0.89   \n",
              "4                  7.23              0.0                 0.77   \n",
              "..                  ...              ...                  ...   \n",
              "115                7.68              0.0                 0.69   \n",
              "116                7.84              0.0                 0.85   \n",
              "117                7.81              0.0                 0.86   \n",
              "118                7.42              0.0                 0.80   \n",
              "119                7.70              0.0                 0.99   \n",
              "\n",
              "     nice_node1_delta  steal_node1_delta  idle_node2_delta  \\\n",
              "0                 0.0                0.0            441.64   \n",
              "1                 0.0                0.0            442.77   \n",
              "2                 0.0                0.0            440.40   \n",
              "3                 0.0                0.0            441.59   \n",
              "4                 0.0                0.0            444.86   \n",
              "..                ...                ...               ...   \n",
              "115               0.0                0.0            443.05   \n",
              "116               0.0                0.0            441.79   \n",
              "117               0.0                0.0            442.36   \n",
              "118               0.0                0.0            443.61   \n",
              "119               0.0                0.0            442.37   \n",
              "\n",
              "     iowait_node2_delta  ...  reads_completed_node2_delta  \\\n",
              "0                  0.44  ...                          0.0   \n",
              "1                  0.43  ...                          0.0   \n",
              "2                  0.47  ...                          0.0   \n",
              "3                  0.39  ...                          0.0   \n",
              "4                  0.41  ...                          0.0   \n",
              "..                  ...  ...                          ...   \n",
              "115                0.43  ...                          0.0   \n",
              "116                0.44  ...                          0.0   \n",
              "117                0.44  ...                          4.0   \n",
              "118                0.43  ...                          0.0   \n",
              "119                0.47  ...                          0.0   \n",
              "\n",
              "     reads_completed_node3_delta  worker1_ens192_delta  worker2_ens192_delta  \\\n",
              "0                            0.0             4419151.0             5929852.0   \n",
              "1                            0.0             4334925.0             5999036.0   \n",
              "2                            0.0             4794640.0             6669287.0   \n",
              "3                            0.0             4614343.0             6190428.0   \n",
              "4                            0.0             4355044.0             5345423.0   \n",
              "..                           ...                   ...                   ...   \n",
              "115                          0.0             4363377.0             5960636.0   \n",
              "116                          0.0             4378204.0             5733231.0   \n",
              "117                          0.0             4493422.0             5586035.0   \n",
              "118                          0.0             4568360.0             5756846.0   \n",
              "119                          0.0             4272198.0             6097997.0   \n",
              "\n",
              "     worker3_ens192_delta  produced_logs  ts-auth-service  ts-travel-service  \\\n",
              "0               9793742.0           2088              485                 27   \n",
              "1               8146510.0           2059              466                 30   \n",
              "2              10555601.0           1985              471                 31   \n",
              "3               8967352.0           1632              376                 21   \n",
              "4               8619615.0           1702              399                 22   \n",
              "..                    ...            ...              ...                ...   \n",
              "115             8111326.0           1690              395                 18   \n",
              "116            13071038.0           1945              467                 25   \n",
              "117             8659602.0           1850              437                 26   \n",
              "118             8719993.0           1879              431                 29   \n",
              "119             8151856.0           1967              457                 26   \n",
              "\n",
              "     ts-ui-dashboard  ts-contacts-service  \n",
              "0                276                  0.0  \n",
              "1                274                  0.0  \n",
              "2                274                  0.0  \n",
              "3                222                  0.0  \n",
              "4                234                  0.0  \n",
              "..               ...                  ...  \n",
              "115              230                  0.0  \n",
              "116              273                  0.0  \n",
              "117              255                  0.0  \n",
              "118              251                  0.0  \n",
              "119              269                  0.0  \n",
              "\n",
              "[120 rows x 50 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1323724c-546f-4241-aa91-4b4908425575\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>idle_node1_delta</th>\n",
              "      <th>iowait_node1_delta</th>\n",
              "      <th>user_node1_delta</th>\n",
              "      <th>system_node1_delta</th>\n",
              "      <th>irq_node1_delta</th>\n",
              "      <th>softirq_node1_delta</th>\n",
              "      <th>nice_node1_delta</th>\n",
              "      <th>steal_node1_delta</th>\n",
              "      <th>idle_node2_delta</th>\n",
              "      <th>iowait_node2_delta</th>\n",
              "      <th>...</th>\n",
              "      <th>reads_completed_node2_delta</th>\n",
              "      <th>reads_completed_node3_delta</th>\n",
              "      <th>worker1_ens192_delta</th>\n",
              "      <th>worker2_ens192_delta</th>\n",
              "      <th>worker3_ens192_delta</th>\n",
              "      <th>produced_logs</th>\n",
              "      <th>ts-auth-service</th>\n",
              "      <th>ts-travel-service</th>\n",
              "      <th>ts-ui-dashboard</th>\n",
              "      <th>ts-contacts-service</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>441.64</td>\n",
              "      <td>0.44</td>\n",
              "      <td>23.35</td>\n",
              "      <td>8.16</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.87</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>441.64</td>\n",
              "      <td>0.44</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4419151.0</td>\n",
              "      <td>5929852.0</td>\n",
              "      <td>9793742.0</td>\n",
              "      <td>2088</td>\n",
              "      <td>485</td>\n",
              "      <td>27</td>\n",
              "      <td>276</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>442.77</td>\n",
              "      <td>0.43</td>\n",
              "      <td>22.99</td>\n",
              "      <td>7.19</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.86</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>442.77</td>\n",
              "      <td>0.43</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4334925.0</td>\n",
              "      <td>5999036.0</td>\n",
              "      <td>8146510.0</td>\n",
              "      <td>2059</td>\n",
              "      <td>466</td>\n",
              "      <td>30</td>\n",
              "      <td>274</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>440.40</td>\n",
              "      <td>0.47</td>\n",
              "      <td>25.34</td>\n",
              "      <td>7.46</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.02</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>440.40</td>\n",
              "      <td>0.47</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4794640.0</td>\n",
              "      <td>6669287.0</td>\n",
              "      <td>10555601.0</td>\n",
              "      <td>1985</td>\n",
              "      <td>471</td>\n",
              "      <td>31</td>\n",
              "      <td>274</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>441.59</td>\n",
              "      <td>0.39</td>\n",
              "      <td>23.98</td>\n",
              "      <td>7.60</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.89</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>441.59</td>\n",
              "      <td>0.39</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4614343.0</td>\n",
              "      <td>6190428.0</td>\n",
              "      <td>8967352.0</td>\n",
              "      <td>1632</td>\n",
              "      <td>376</td>\n",
              "      <td>21</td>\n",
              "      <td>222</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>444.86</td>\n",
              "      <td>0.41</td>\n",
              "      <td>21.26</td>\n",
              "      <td>7.23</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.77</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>444.86</td>\n",
              "      <td>0.41</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4355044.0</td>\n",
              "      <td>5345423.0</td>\n",
              "      <td>8619615.0</td>\n",
              "      <td>1702</td>\n",
              "      <td>399</td>\n",
              "      <td>22</td>\n",
              "      <td>234</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>115</th>\n",
              "      <td>443.05</td>\n",
              "      <td>0.43</td>\n",
              "      <td>22.67</td>\n",
              "      <td>7.68</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.69</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>443.05</td>\n",
              "      <td>0.43</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4363377.0</td>\n",
              "      <td>5960636.0</td>\n",
              "      <td>8111326.0</td>\n",
              "      <td>1690</td>\n",
              "      <td>395</td>\n",
              "      <td>18</td>\n",
              "      <td>230</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>116</th>\n",
              "      <td>441.79</td>\n",
              "      <td>0.44</td>\n",
              "      <td>23.15</td>\n",
              "      <td>7.84</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.85</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>441.79</td>\n",
              "      <td>0.44</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4378204.0</td>\n",
              "      <td>5733231.0</td>\n",
              "      <td>13071038.0</td>\n",
              "      <td>1945</td>\n",
              "      <td>467</td>\n",
              "      <td>25</td>\n",
              "      <td>273</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>117</th>\n",
              "      <td>442.36</td>\n",
              "      <td>0.44</td>\n",
              "      <td>22.97</td>\n",
              "      <td>7.81</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.86</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>442.36</td>\n",
              "      <td>0.44</td>\n",
              "      <td>...</td>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4493422.0</td>\n",
              "      <td>5586035.0</td>\n",
              "      <td>8659602.0</td>\n",
              "      <td>1850</td>\n",
              "      <td>437</td>\n",
              "      <td>26</td>\n",
              "      <td>255</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>118</th>\n",
              "      <td>443.61</td>\n",
              "      <td>0.43</td>\n",
              "      <td>21.65</td>\n",
              "      <td>7.42</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.80</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>443.61</td>\n",
              "      <td>0.43</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4568360.0</td>\n",
              "      <td>5756846.0</td>\n",
              "      <td>8719993.0</td>\n",
              "      <td>1879</td>\n",
              "      <td>431</td>\n",
              "      <td>29</td>\n",
              "      <td>251</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>119</th>\n",
              "      <td>442.37</td>\n",
              "      <td>0.47</td>\n",
              "      <td>23.06</td>\n",
              "      <td>7.70</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.99</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>442.37</td>\n",
              "      <td>0.47</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>4272198.0</td>\n",
              "      <td>6097997.0</td>\n",
              "      <td>8151856.0</td>\n",
              "      <td>1967</td>\n",
              "      <td>457</td>\n",
              "      <td>26</td>\n",
              "      <td>269</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>120 rows × 50 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1323724c-546f-4241-aa91-4b4908425575')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-1323724c-546f-4241-aa91-4b4908425575 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-1323724c-546f-4241-aa91-4b4908425575');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 153
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_data_scaled = scaler.fit_transform(test_lstm)"
      ],
      "metadata": {
        "id": "_442ETBLq1HA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "from sklearn.metrics import mean_squared_error\n",
        "\n",
        "# Step 1: Prepare your dataset\n",
        "# Assuming you have a numpy array 'X_train' containing normal instances with shape (num_samples, num_features)\n",
        "\n",
        "# Step 2: Sequence Representation\n",
        "# Convert your input data into sequential representations suitable for LSTM input\n",
        "sequence_length = 10  # Length of each sequence\n",
        "num_features = 50  # Number of features in the dataset\n",
        "X_train_sequences = []\n",
        "for i in range(len(train_data_scaled) - sequence_length):\n",
        "    sequence = train_data_scaled[i:i + sequence_length]\n",
        "    X_train_sequences.append(sequence)\n",
        "X_train_sequences = np.array(X_train_sequences)\n",
        "\n",
        "X_val1_sequences = []\n",
        "for i in range(len(val_data1_scaled) - sequence_length):\n",
        "    sequence = val_data1_scaled[i:i + sequence_length]\n",
        "    X_val1_sequences.append(sequence)\n",
        "X_val1_sequences = np.array(X_val1_sequences)\n",
        "\n",
        "# Step 3: Model Architecture\n",
        "model = tf.keras.Sequential()\n",
        "model.add(tf.keras.layers.LSTM(32, input_shape=(sequence_length, num_features)))\n",
        "model.add(tf.keras.layers.Dense(1, activation='sigmoid'))\n",
        "\n",
        "# Step 4: Training\n",
        "model.compile(loss='mse', optimizer='adam')\n",
        "model.fit(X_train_sequences, np.ones(len(X_train_sequences)), epochs=100, validation_data=(X_val1_sequences, np.ones(len(X_val1_sequences))))\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "efszU1-DlZ5m",
        "outputId": "1268a48a-1d31-4368-af86-57bdafb8f9dc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "142/142 [==============================] - 4s 11ms/step - loss: 0.0033 - val_loss: 0.0111\n",
            "Epoch 2/100\n",
            "142/142 [==============================] - 1s 9ms/step - loss: 1.1306e-04 - val_loss: 0.0079\n",
            "Epoch 3/100\n",
            "142/142 [==============================] - 2s 11ms/step - loss: 5.2614e-05 - val_loss: 0.0062\n",
            "Epoch 4/100\n",
            "142/142 [==============================] - 2s 11ms/step - loss: 3.1117e-05 - val_loss: 0.0052\n",
            "Epoch 5/100\n",
            "142/142 [==============================] - 1s 9ms/step - loss: 2.0815e-05 - val_loss: 0.0044\n",
            "Epoch 6/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 1.5107e-05 - val_loss: 0.0039\n",
            "Epoch 7/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 1.1616e-05 - val_loss: 0.0035\n",
            "Epoch 8/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 9.2824e-06 - val_loss: 0.0031\n",
            "Epoch 9/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 7.6139e-06 - val_loss: 0.0028\n",
            "Epoch 10/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 6.3657e-06 - val_loss: 0.0026\n",
            "Epoch 11/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 5.4018e-06 - val_loss: 0.0024\n",
            "Epoch 12/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 4.6399e-06 - val_loss: 0.0022\n",
            "Epoch 13/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 4.0264e-06 - val_loss: 0.0020\n",
            "Epoch 14/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 3.5248e-06 - val_loss: 0.0019\n",
            "Epoch 15/100\n",
            "142/142 [==============================] - 2s 11ms/step - loss: 3.1092e-06 - val_loss: 0.0018\n",
            "Epoch 16/100\n",
            "142/142 [==============================] - 3s 20ms/step - loss: 2.7609e-06 - val_loss: 0.0017\n",
            "Epoch 17/100\n",
            "142/142 [==============================] - 3s 18ms/step - loss: 2.4657e-06 - val_loss: 0.0016\n",
            "Epoch 18/100\n",
            "142/142 [==============================] - 2s 16ms/step - loss: 2.2132e-06 - val_loss: 0.0015\n",
            "Epoch 19/100\n",
            "142/142 [==============================] - 2s 15ms/step - loss: 1.9953e-06 - val_loss: 0.0014\n",
            "Epoch 20/100\n",
            "142/142 [==============================] - 2s 13ms/step - loss: 1.8059e-06 - val_loss: 0.0013\n",
            "Epoch 21/100\n",
            "142/142 [==============================] - 1s 9ms/step - loss: 1.6400e-06 - val_loss: 0.0012\n",
            "Epoch 22/100\n",
            "142/142 [==============================] - 2s 13ms/step - loss: 1.4937e-06 - val_loss: 0.0012\n",
            "Epoch 23/100\n",
            "142/142 [==============================] - 3s 20ms/step - loss: 1.3641e-06 - val_loss: 0.0011\n",
            "Epoch 24/100\n",
            "142/142 [==============================] - 3s 18ms/step - loss: 1.2486e-06 - val_loss: 0.0010\n",
            "Epoch 25/100\n",
            "142/142 [==============================] - 2s 15ms/step - loss: 1.1452e-06 - val_loss: 9.7984e-04\n",
            "Epoch 26/100\n",
            "142/142 [==============================] - 2s 17ms/step - loss: 1.0523e-06 - val_loss: 9.2853e-04\n",
            "Epoch 27/100\n",
            "142/142 [==============================] - 2s 14ms/step - loss: 9.6841e-07 - val_loss: 8.8044e-04\n",
            "Epoch 28/100\n",
            "142/142 [==============================] - 2s 14ms/step - loss: 8.9251e-07 - val_loss: 8.3530e-04\n",
            "Epoch 29/100\n",
            "142/142 [==============================] - 3s 23ms/step - loss: 8.2360e-07 - val_loss: 7.9286e-04\n",
            "Epoch 30/100\n",
            "142/142 [==============================] - 3s 24ms/step - loss: 7.6088e-07 - val_loss: 7.5290e-04\n",
            "Epoch 31/100\n",
            "142/142 [==============================] - 2s 16ms/step - loss: 7.0365e-07 - val_loss: 7.1524e-04\n",
            "Epoch 32/100\n",
            "142/142 [==============================] - 1s 9ms/step - loss: 6.5131e-07 - val_loss: 6.7969e-04\n",
            "Epoch 33/100\n",
            "142/142 [==============================] - 2s 12ms/step - loss: 6.0335e-07 - val_loss: 6.4610e-04\n",
            "Epoch 34/100\n",
            "142/142 [==============================] - 1s 10ms/step - loss: 5.5933e-07 - val_loss: 6.1435e-04\n",
            "Epoch 35/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 5.1887e-07 - val_loss: 5.8429e-04\n",
            "Epoch 36/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 4.8161e-07 - val_loss: 5.5586e-04\n",
            "Epoch 37/100\n",
            "142/142 [==============================] - 2s 11ms/step - loss: 4.4727e-07 - val_loss: 5.2891e-04\n",
            "Epoch 38/100\n",
            "142/142 [==============================] - 2s 11ms/step - loss: 4.1558e-07 - val_loss: 5.0338e-04\n",
            "Epoch 39/100\n",
            "142/142 [==============================] - 2s 11ms/step - loss: 3.8629e-07 - val_loss: 4.7916e-04\n",
            "Epoch 40/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 3.5922e-07 - val_loss: 4.5618e-04\n",
            "Epoch 41/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 3.3415e-07 - val_loss: 4.3436e-04\n",
            "Epoch 42/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 3.1093e-07 - val_loss: 4.1365e-04\n",
            "Epoch 43/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 2.8941e-07 - val_loss: 3.9397e-04\n",
            "Epoch 44/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 2.6945e-07 - val_loss: 3.7527e-04\n",
            "Epoch 45/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 2.5092e-07 - val_loss: 3.5750e-04\n",
            "Epoch 46/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 2.3371e-07 - val_loss: 3.4060e-04\n",
            "Epoch 47/100\n",
            "142/142 [==============================] - 2s 12ms/step - loss: 2.1773e-07 - val_loss: 3.2454e-04\n",
            "Epoch 48/100\n",
            "142/142 [==============================] - 3s 18ms/step - loss: 2.0287e-07 - val_loss: 3.0927e-04\n",
            "Epoch 49/100\n",
            "142/142 [==============================] - 5s 33ms/step - loss: 1.8906e-07 - val_loss: 2.9475e-04\n",
            "Epoch 50/100\n",
            "142/142 [==============================] - 2s 14ms/step - loss: 1.7621e-07 - val_loss: 2.8093e-04\n",
            "Epoch 51/100\n",
            "142/142 [==============================] - 2s 15ms/step - loss: 1.6425e-07 - val_loss: 2.6779e-04\n",
            "Epoch 52/100\n",
            "142/142 [==============================] - 2s 17ms/step - loss: 1.5312e-07 - val_loss: 2.5529e-04\n",
            "Epoch 53/100\n",
            "142/142 [==============================] - 2s 13ms/step - loss: 1.4277e-07 - val_loss: 2.4339e-04\n",
            "Epoch 54/100\n",
            "142/142 [==============================] - 2s 16ms/step - loss: 1.3312e-07 - val_loss: 2.3206e-04\n",
            "Epoch 55/100\n",
            "142/142 [==============================] - 3s 22ms/step - loss: 1.2413e-07 - val_loss: 2.2129e-04\n",
            "Epoch 56/100\n",
            "142/142 [==============================] - 2s 17ms/step - loss: 1.1576e-07 - val_loss: 2.1103e-04\n",
            "Epoch 57/100\n",
            "142/142 [==============================] - 2s 13ms/step - loss: 1.0797e-07 - val_loss: 2.0128e-04\n",
            "Epoch 58/100\n",
            "142/142 [==============================] - 2s 12ms/step - loss: 1.0070e-07 - val_loss: 1.9199e-04\n",
            "Epoch 59/100\n",
            "142/142 [==============================] - 2s 11ms/step - loss: 9.3926e-08 - val_loss: 1.8315e-04\n",
            "Epoch 60/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 8.7614e-08 - val_loss: 1.7473e-04\n",
            "Epoch 61/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 8.1730e-08 - val_loss: 1.6672e-04\n",
            "Epoch 62/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 7.6244e-08 - val_loss: 1.5910e-04\n",
            "Epoch 63/100\n",
            "142/142 [==============================] - 1s 9ms/step - loss: 7.1129e-08 - val_loss: 1.5184e-04\n",
            "Epoch 64/100\n",
            "142/142 [==============================] - 2s 12ms/step - loss: 6.6360e-08 - val_loss: 1.4493e-04\n",
            "Epoch 65/100\n",
            "142/142 [==============================] - 2s 13ms/step - loss: 6.1913e-08 - val_loss: 1.3835e-04\n",
            "Epoch 66/100\n",
            "142/142 [==============================] - 2s 16ms/step - loss: 5.7765e-08 - val_loss: 1.3208e-04\n",
            "Epoch 67/100\n",
            "142/142 [==============================] - 2s 16ms/step - loss: 5.3897e-08 - val_loss: 1.2612e-04\n",
            "Epoch 68/100\n",
            "142/142 [==============================] - 2s 11ms/step - loss: 5.0290e-08 - val_loss: 1.2044e-04\n",
            "Epoch 69/100\n",
            "142/142 [==============================] - 2s 13ms/step - loss: 4.6926e-08 - val_loss: 1.1504e-04\n",
            "Epoch 70/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 4.3788e-08 - val_loss: 1.0990e-04\n",
            "Epoch 71/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 4.0861e-08 - val_loss: 1.0500e-04\n",
            "Epoch 72/100\n",
            "142/142 [==============================] - 2s 11ms/step - loss: 3.8131e-08 - val_loss: 1.0033e-04\n",
            "Epoch 73/100\n",
            "142/142 [==============================] - 2s 12ms/step - loss: 3.5584e-08 - val_loss: 9.5892e-05\n",
            "Epoch 74/100\n",
            "142/142 [==============================] - 2s 11ms/step - loss: 3.3209e-08 - val_loss: 9.1665e-05\n",
            "Epoch 75/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 3.0993e-08 - val_loss: 8.7640e-05\n",
            "Epoch 76/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 2.8926e-08 - val_loss: 8.3808e-05\n",
            "Epoch 77/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 2.6999e-08 - val_loss: 8.0158e-05\n",
            "Epoch 78/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 2.5200e-08 - val_loss: 7.6683e-05\n",
            "Epoch 79/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 2.3522e-08 - val_loss: 7.3374e-05\n",
            "Epoch 80/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 2.1957e-08 - val_loss: 7.0223e-05\n",
            "Epoch 81/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 2.0497e-08 - val_loss: 6.7222e-05\n",
            "Epoch 82/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 1.9136e-08 - val_loss: 6.4364e-05\n",
            "Epoch 83/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 1.7866e-08 - val_loss: 6.1641e-05\n",
            "Epoch 84/100\n",
            "142/142 [==============================] - 2s 12ms/step - loss: 1.6681e-08 - val_loss: 5.9047e-05\n",
            "Epoch 85/100\n",
            "142/142 [==============================] - 2s 14ms/step - loss: 1.5575e-08 - val_loss: 5.6575e-05\n",
            "Epoch 86/100\n",
            "142/142 [==============================] - 2s 14ms/step - loss: 1.4544e-08 - val_loss: 5.4222e-05\n",
            "Epoch 87/100\n",
            "142/142 [==============================] - 2s 14ms/step - loss: 1.3582e-08 - val_loss: 5.1978e-05\n",
            "Epoch 88/100\n",
            "142/142 [==============================] - 2s 12ms/step - loss: 1.2685e-08 - val_loss: 4.9840e-05\n",
            "Epoch 89/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 1.1848e-08 - val_loss: 4.7802e-05\n",
            "Epoch 90/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 1.1068e-08 - val_loss: 4.5860e-05\n",
            "Epoch 91/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 1.0340e-08 - val_loss: 4.4009e-05\n",
            "Epoch 92/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 9.6603e-09 - val_loss: 4.2243e-05\n",
            "Epoch 93/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 9.0266e-09 - val_loss: 4.0559e-05\n",
            "Epoch 94/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 8.4357e-09 - val_loss: 3.8953e-05\n",
            "Epoch 95/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 7.8843e-09 - val_loss: 3.7421e-05\n",
            "Epoch 96/100\n",
            "142/142 [==============================] - 1s 7ms/step - loss: 7.3699e-09 - val_loss: 3.5959e-05\n",
            "Epoch 97/100\n",
            "142/142 [==============================] - 1s 8ms/step - loss: 6.8903e-09 - val_loss: 3.4565e-05\n",
            "Epoch 98/100\n",
            "142/142 [==============================] - 2s 12ms/step - loss: 6.4429e-09 - val_loss: 3.3234e-05\n",
            "Epoch 99/100\n",
            "142/142 [==============================] - 2s 12ms/step - loss: 6.0255e-09 - val_loss: 3.1964e-05\n",
            "Epoch 100/100\n",
            "142/142 [==============================] - 2s 15ms/step - loss: 5.6361e-09 - val_loss: 3.0751e-05\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7ff22b896080>"
            ]
          },
          "metadata": {},
          "execution_count": 158
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(val_data_scaled)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "etCOl0HI_abj",
        "outputId": "d4c38897-bb3a-451d-ad85-6e0d9cfb55ff"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1134"
            ]
          },
          "metadata": {},
          "execution_count": 119
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_val2_sequences = []\n",
        "for i in range(len(val_data2_scaled) - sequence_length):\n",
        "    sequence = val_data2_scaled[i:i + sequence_length]\n",
        "    X_val2_sequences.append(sequence)\n",
        "X_val2_sequences = np.array(X_val2_sequences)\n",
        "# Step 5: Threshold Determination\n",
        "X_val_pred = model.predict(X_val2_sequences)\n",
        "val_mse = mean_squared_error(np.ones(len(X_val2_sequences)),X_val_pred)\n",
        "# Set the threshold based on a statistical measure (e.g., mean plus a certain number of standard deviations)\n",
        "#threshold = np.mean(val_mse) + 2 * np.std(val_mse)\n",
        "#threshold = np.mean(val_mse) + (np.std(val_mse)/np.sqrt(len(val_data_scaled)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Wu8mw-XOnSIi",
        "outputId": "f317bd39-64aa-46a3-eed7-4921c14db243"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "18/18 [==============================] - 0s 3ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.scatter(val_mse)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 183
        },
        "id": "_4a0Cb_2Z3RI",
        "outputId": "00f17c92-35f5-4756-b165-1ba846e419f3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "TypeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-183-620a876fd8f1>\u001b[0m in \u001b[0;36m<cell line: 1>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mscatter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mval_mse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshow\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mTypeError\u001b[0m: scatter() missing 1 required positional argument: 'y'"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.mean(val_mse) + (np.std(val_mse)/np.sqrt(len(val_data2_scaled)))"
      ],
      "metadata": {
        "id": "aGN2olA-ZIBK"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.mean(val_mse) + 2*np.std(val_mse)"
      ],
      "metadata": {
        "id": "vum2MeUxZuyc"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = np.percentile(val_mse, 95)"
      ],
      "metadata": {
        "id": "O4DZm7U0aMao"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "threshold"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bpdv2qFOoDOR",
        "outputId": "2fa0b348-4828-4eb3-a3ea-966a1689b74d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "4.872710816600145e-05"
            ]
          },
          "metadata": {},
          "execution_count": 187
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Step 6: Anomaly Detection\n",
        "# Assuming you have a numpy array 'X_test' containing new instances\n",
        "X_test_sequences = []\n",
        "for i in range(len(test_data_scaled) - sequence_length):\n",
        "    sequence = test_data_scaled[i:i + sequence_length]\n",
        "    X_test_sequences.append(sequence)\n",
        "X_test_sequences = np.array(X_test_sequences)\n",
        "\n",
        "X_test_pred = model.predict(X_test_sequences)\n",
        "test_mse = mean_squared_error(np.ones(len(X_test_sequences)), X_test_pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "y1d7hutMl0H7",
        "outputId": "4d76cb77-e2b0-4e37-d102-bbcad39d5bad"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4/4 [==============================] - 0s 4ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_mse = np.sqrt(X_test_pred - test_data_scaled[10:])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZmVzf0GrmzZf",
        "outputId": "4108eede-7d51-41ed-a410-bbd11e270298"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-164-bd2ba03d4b12>:1: RuntimeWarning: invalid value encountered in sqrt\n",
            "  test_mse = np.sqrt(X_test_pred - test_data_scaled[10:])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_mse"
      ],
      "metadata": {
        "id": "BvNyIWmJmpWl",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "80cf4806-2a0b-4fb6-e688-f5cddd375e97"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.49306877, 0.98524934, 0.81396839, ..., 0.77420361, 0.54892838,\n",
              "        0.98711257],\n",
              "       [0.71485765, 0.98793642, 0.65887476, ..., 0.9255998 , 0.08202866,\n",
              "        0.99082541],\n",
              "       [0.97956844, 0.98972469,        nan, ..., 0.93124282, 0.3701019 ,\n",
              "        0.99609899],\n",
              "       ...,\n",
              "       [0.40111816, 0.98191696, 0.82072114, ..., 0.72796123, 0.9799482 ,\n",
              "        0.98357895],\n",
              "       [0.33432822, 0.97970519, 0.86341836, ..., 0.65683375, 0.97769473,\n",
              "        0.98116287],\n",
              "       [0.39739777, 0.97995924, 0.81592947, ..., 0.72616234, 0.97801251,\n",
              "        0.98224831]])"
            ]
          },
          "metadata": {},
          "execution_count": 165
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "predictions=[]\n",
        "for x in test_mse:\n",
        "  if np.mean(x) > threshold:\n",
        "    predictions.append(1)\n",
        "  else:\n",
        "    predictions.append(0)\n",
        "predictions"
      ],
      "metadata": {
        "id": "Ymw3j_OgmTjj",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f6e11ef1-da87-4801-c3da-0daed42ae9cd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 0,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 0,\n",
              " 1,\n",
              " 1,\n",
              " 1]"
            ]
          },
          "metadata": {},
          "execution_count": 188
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(predictions)"
      ],
      "metadata": {
        "id": "zM_qVLUPsIwU",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c0c59c1e-59f9-4063-f4b3-0773fa78a8af"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "110"
            ]
          },
          "metadata": {},
          "execution_count": 189
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(gt[:110], predictions, target_names=['normal', 'attack']))"
      ],
      "metadata": {
        "id": "Wles73WkmUku",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6039b74b-d7f8-4b51-fbd7-c608a5c20ca0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      normal       0.56      0.23      0.33        61\n",
            "      attack       0.45      0.78      0.57        49\n",
            "\n",
            "    accuracy                           0.47       110\n",
            "   macro avg       0.50      0.50      0.45       110\n",
            "weighted avg       0.51      0.47      0.43       110\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "mq9Zj9I0rpJP"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}