HAR_LSTM HYPERPARATEMER TUNING.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Exercise : HAR_LSTM HYPERPARATEMER TUNING\n",
    "## Step by step Process \n",
    "- Lets start as we know This dataset is collected from 30 persons(referred as subjects in this dataset), performing different activities with a smartphone to their waists. The data is recorded with the help of sensors (accelerometer and Gyroscope) in that smartphone. This experiment was video recorded to label the data manually.\n",
    "- So after performing lots of EDA and visulization on dataset we have performed both Classical ML models and Deep learning RNN models and we know we haved two types of data:  \n",
    "    1. Expert Engg features - perform Classical ML models  \n",
    "    2. Raw time series data - perform Deep learning RNN models \n",
    "- After doing all above in this in this we will performs some hyperparameter tuning on our deep learning models.\n",
    "- Lets Start working :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing Libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "from sklearn.metrics import confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Activities are the class labels\n",
    "# It is a 6 class classification\n",
    "ACTIVITIES = {\n",
    "    0: 'WALKING',\n",
    "    1: 'WALKING_UPSTAIRS',\n",
    "    2: 'WALKING_DOWNSTAIRS',\n",
    "    3: 'SITTING',\n",
    "    4: 'STANDING',\n",
    "    5: 'LAYING',\n",
    "}\n",
    "\n",
    "# Utility function to print the confusion matrix\n",
    "def confusion_matrix(Y_true, Y_pred):\n",
    "    Y_true = pd.Series([ACTIVITIES[y] for y in np.argmax(Y_true, axis=1)])\n",
    "    Y_pred = pd.Series([ACTIVITIES[y] for y in np.argmax(Y_pred, axis=1)])\n",
    "\n",
    "    return pd.crosstab(Y_true, Y_pred, rownames=['True'], colnames=['Pred'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data directory\n",
    "DATADIR = 'UCI_HAR_Dataset'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Raw data signals\n",
    "# Signals are from Accelerometer and Gyroscope\n",
    "# The signals are in x,y,z directions\n",
    "# Sensor signals are filtered to have only body acceleration\n",
    "# excluding the acceleration due to gravity\n",
    "# Triaxial acceleration from the accelerometer is total acceleration\n",
    "SIGNALS = [\n",
    "    \"body_acc_x\",\n",
    "    \"body_acc_y\",\n",
    "    \"body_acc_z\",\n",
    "    \"body_gyro_x\",\n",
    "    \"body_gyro_y\",\n",
    "    \"body_gyro_z\",\n",
    "    \"total_acc_x\",\n",
    "    \"total_acc_y\",\n",
    "    \"total_acc_z\"\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Utility function to read the data from csv file\n",
    "def _read_csv(filename):\n",
    "    return pd.read_csv(filename, delim_whitespace=True, header=None)\n",
    "\n",
    "# Utility function to load the load\n",
    "def load_signals(subset):\n",
    "    signals_data = []\n",
    "\n",
    "    for signal in SIGNALS:\n",
    "        filename = f'UCI_HAR_Dataset/{subset}/Inertial Signals/{signal}_{subset}.txt'\n",
    "        signals_data.append(\n",
    "            _read_csv(filename).as_matrix()\n",
    "        ) \n",
    "\n",
    "    # Transpose is used to change the dimensionality of the output,\n",
    "    # aggregating the signals by combination of sample/timestep.\n",
    "    # Resultant shape is (7352 train/2947 test samples, 128 timesteps, 9 signals)\n",
    "    return np.transpose(signals_data, (1, 2, 0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def load_y(subset):\n",
    "    \"\"\"\n",
    "    The objective that we are trying to predict is a integer, from 1 to 6,\n",
    "    that represents a human activity. We return a binary representation of \n",
    "    every sample objective as a 6 bits vector using One Hot Encoding\n",
    "    (https://pandas.pydata.org/pandas-docs/stable/generated/pandas.get_dummies.html)\n",
    "    \"\"\"\n",
    "    filename = f'UCI_HAR_Dataset/{subset}/y_{subset}.txt'\n",
    "    y = _read_csv(filename)[0]\n",
    "\n",
    "    return pd.get_dummies(y).as_matrix()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_data():\n",
    "    \"\"\"\n",
    "    Obtain the dataset from multiple files.\n",
    "    Returns: X_train, X_test, y_train, y_test\n",
    "    \"\"\"\n",
    "    X_train, X_test = load_signals('train'), load_signals('test')\n",
    "    y_train, y_test = load_y('train'), load_y('test')\n",
    "\n",
    "    return X_train, X_test, y_train, y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# utility function\n",
    "def plt_dynamic(x, vy, ty):\n",
    "    plt.figure(figsize=(10,5))\n",
    "    plt.plot(x, vy, 'b', label=\"Validation Loss\")\n",
    "    plt.plot(x, ty, 'r', label=\"Train Loss\")\n",
    "    plt.xlabel('Epochs') \n",
    "    plt.ylabel('Crossentropy Loss')\n",
    "    plt.title('Crossentropy Loss VS Epochs')\n",
    "    plt.legend()\n",
    "    plt.grid()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\nisha\\Anaconda3\\lib\\site-packages\\h5py\\__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n"
     ]
    }
   ],
   "source": [
    "# Importing tensorflow\n",
    "np.random.seed(42)\n",
    "import tensorflow as tf\n",
    "tf.set_random_seed(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Configuring a session\n",
    "session_conf = tf.ConfigProto(\n",
    "    intra_op_parallelism_threads=1,\n",
    "    inter_op_parallelism_threads=1\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "# Import Keras\n",
    "from keras import backend as K\n",
    "sess = tf.Session(graph=tf.get_default_graph(), config=session_conf)\n",
    "K.set_session(sess)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing libraries\n",
    "from keras.models import Sequential\n",
    "from keras.layers import LSTM\n",
    "from keras.layers.core import Dense, Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initializing parameters\n",
    "epochs = 30\n",
    "batch_size = 16\n",
    "n_hidden = 32"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Utility function to count the number of classes\n",
    "def _count_classes(y):\n",
    "    return len(set([tuple(category) for category in y]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\nisha\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:12: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
      "  if sys.path[0] == '':\n"
     ]
    }
   ],
   "source": [
    "# Loading the train and test data\n",
    "X_train, X_test, Y_train, Y_test = load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "128\n",
      "9\n",
      "7352\n"
     ]
    }
   ],
   "source": [
    "timesteps = len(X_train[0])\n",
    "input_dim = len(X_train[0][0])\n",
    "n_classes = _count_classes(Y_train)\n",
    "\n",
    "print(timesteps)\n",
    "print(input_dim)\n",
    "print(len(X_train))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Defining the Architecture of LSTM"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Model having 1 LSTM layer with 32 LSTM Units"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_1 (LSTM)                (None, 32)                5376      \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 32)                0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 6)                 198       \n",
      "=================================================================\n",
      "Total params: 5,574\n",
      "Trainable params: 5,574\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Train on 7352 samples, validate on 2947 samples\n",
      "Epoch 1/30\n",
      "7352/7352 [==============================] - 39s 5ms/step - loss: 1.3139 - acc: 0.4358 - val_loss: 1.1352 - val_acc: 0.4700\n",
      "Epoch 2/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.9788 - acc: 0.5773 - val_loss: 0.9513 - val_acc: 0.5884\n",
      "Epoch 3/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.7977 - acc: 0.6457 - val_loss: 0.8343 - val_acc: 0.6013\n",
      "Epoch 4/30\n",
      "7352/7352 [==============================] - 39s 5ms/step - loss: 0.6989 - acc: 0.6582 - val_loss: 0.7532 - val_acc: 0.6098\n",
      "Epoch 5/30\n",
      "7352/7352 [==============================] - 36s 5ms/step - loss: 0.6359 - acc: 0.6797 - val_loss: 0.7335 - val_acc: 0.6183\n",
      "Epoch 6/30\n",
      "7352/7352 [==============================] - 42s 6ms/step - loss: 0.5819 - acc: 0.6865 - val_loss: 0.8786 - val_acc: 0.6098\n",
      "Epoch 7/30\n",
      "7352/7352 [==============================] - 43s 6ms/step - loss: 0.5676 - acc: 0.7058 - val_loss: 0.8191 - val_acc: 0.6132\n",
      "Epoch 8/30\n",
      "7352/7352 [==============================] - 42s 6ms/step - loss: 0.5583 - acc: 0.7217 - val_loss: 0.6639 - val_acc: 0.7190\n",
      "Epoch 9/30\n",
      "7352/7352 [==============================] - 39s 5ms/step - loss: 0.5386 - acc: 0.7557 - val_loss: 0.6388 - val_acc: 0.7167\n",
      "Epoch 10/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.4804 - acc: 0.7911 - val_loss: 0.5077 - val_acc: 0.7509\n",
      "Epoch 11/30\n",
      "7352/7352 [==============================] - 38s 5ms/step - loss: 0.4320 - acc: 0.8052 - val_loss: 0.5143 - val_acc: 0.7418\n",
      "Epoch 12/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.4279 - acc: 0.8062 - val_loss: 0.4951 - val_acc: 0.7472\n",
      "Epoch 13/30\n",
      "7352/7352 [==============================] - 39s 5ms/step - loss: 0.3911 - acc: 0.8130 - val_loss: 0.5606 - val_acc: 0.7516\n",
      "Epoch 14/30\n",
      "7352/7352 [==============================] - 39s 5ms/step - loss: 0.3898 - acc: 0.8313 - val_loss: 0.4518 - val_acc: 0.8137\n",
      "Epoch 15/30\n",
      "7352/7352 [==============================] - 42s 6ms/step - loss: 0.3308 - acc: 0.8942 - val_loss: 0.4732 - val_acc: 0.8633\n",
      "Epoch 16/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.2891 - acc: 0.9176 - val_loss: 0.3794 - val_acc: 0.8765\n",
      "Epoch 17/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.2660 - acc: 0.9246 - val_loss: 0.5082 - val_acc: 0.8660\n",
      "Epoch 18/30\n",
      "7352/7352 [==============================] - 38s 5ms/step - loss: 0.2538 - acc: 0.9251 - val_loss: 0.4772 - val_acc: 0.8806\n",
      "Epoch 19/30\n",
      "7352/7352 [==============================] - 44s 6ms/step - loss: 0.2502 - acc: 0.9312 - val_loss: 0.7013 - val_acc: 0.8307\n",
      "Epoch 20/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.1980 - acc: 0.9382 - val_loss: 0.3988 - val_acc: 0.8890\n",
      "Epoch 21/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.2018 - acc: 0.9372 - val_loss: 1.7682 - val_acc: 0.7075\n",
      "Epoch 22/30\n",
      "7352/7352 [==============================] - 36s 5ms/step - loss: 0.2455 - acc: 0.9310 - val_loss: 0.5812 - val_acc: 0.8687\n",
      "Epoch 23/30\n",
      "7352/7352 [==============================] - 38s 5ms/step - loss: 0.2194 - acc: 0.9329 - val_loss: 0.6468 - val_acc: 0.8744\n",
      "Epoch 24/30\n",
      "7352/7352 [==============================] - 39s 5ms/step - loss: 0.2282 - acc: 0.9304 - val_loss: 0.4721 - val_acc: 0.8741\n",
      "Epoch 25/30\n",
      "7352/7352 [==============================] - 42s 6ms/step - loss: 0.2166 - acc: 0.9359 - val_loss: 0.4131 - val_acc: 0.8938\n",
      "Epoch 26/30\n",
      "7352/7352 [==============================] - 45s 6ms/step - loss: 0.2173 - acc: 0.9350 - val_loss: 0.4841 - val_acc: 0.8887\n",
      "Epoch 27/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.2224 - acc: 0.9353 - val_loss: 0.3590 - val_acc: 0.8935\n",
      "Epoch 28/30\n",
      "7352/7352 [==============================] - 45s 6ms/step - loss: 0.1961 - acc: 0.9385 - val_loss: 0.5297 - val_acc: 0.8802\n",
      "Epoch 29/30\n",
      "7352/7352 [==============================] - 48s 7ms/step - loss: 0.1876 - acc: 0.9416 - val_loss: 0.4324 - val_acc: 0.8924\n",
      "Epoch 30/30\n",
      "7352/7352 [==============================] - 48s 7ms/step - loss: 0.1999 - acc: 0.9411 - val_loss: 0.4883 - val_acc: 0.8829\n"
     ]
    }
   ],
   "source": [
    "# Initiliazing the sequential model\n",
    "model1 = Sequential()\n",
    "# Configuring the parameters\n",
    "model1.add(LSTM(n_hidden, input_shape=(timesteps, input_dim)))\n",
    "# Adding a dropout layer\n",
    "model1.add(Dropout(0.5))\n",
    "# Adding a dense output layer with sigmoid activation\n",
    "model1.add(Dense(n_classes, activation='sigmoid'))\n",
    "model1.summary()\n",
    "\n",
    "# Compiling the model\n",
    "model1.compile(loss='categorical_crossentropy',optimizer='rmsprop',metrics=['accuracy'])\n",
    "\n",
    "# Training the model\n",
    "history1 = model1.fit(X_train, Y_train, batch_size=batch_size,validation_data=(X_test, Y_test),epochs=epochs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Score: 0.488270\n",
      "Test Accuracy: 88.293180%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Final evaluation of the model\n",
    "scores = model1.evaluate(X_test, Y_test, verbose=0)\n",
    "print(\"Test Score: %f\" % (scores[0]))\n",
    "print(\"Test Accuracy: %f%%\" % (scores[1]*100))\n",
    "\n",
    "# Confusion Matrix\n",
    "Y_true = pd.Series([ACTIVITIES[y] for y in np.argmax(Y_test, axis=1)])\n",
    "Y_predictions = pd.Series([ACTIVITIES[y] for y in np.argmax(model1.predict(X_test), axis=1)])\n",
    "\n",
    "# Code for drawing seaborn heatmaps\n",
    "class_names = ['laying','sitting','standing','walking','walking_downstairs','walking_upstairs']\n",
    "df_heatmap = pd.DataFrame(confusion_matrix(Y_true, Y_predictions), index=class_names, columns=class_names )\n",
    "fig = plt.figure(figsize=(10,7))\n",
    "heatmap = sns.heatmap(df_heatmap, annot=True, fmt=\"d\")\n",
    "\n",
    "# Setting tick labels for heatmap\n",
    "heatmap.yaxis.set_ticklabels(heatmap.yaxis.get_ticklabels(), rotation=0, ha='right', fontsize=14)\n",
    "heatmap.xaxis.set_ticklabels(heatmap.xaxis.get_ticklabels(), rotation=90, ha='right', fontsize=14)\n",
    "plt.ylabel('True label',size=18)\n",
    "plt.xlabel('Predicted label',size=18)\n",
    "plt.title(\"Confusion Matrix\\n\",size=24)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 88.29%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Final evaluation of the model\n",
    "scores = model1.evaluate(X_test, Y_test, verbose=0)\n",
    "print(\"Accuracy: %.2f%%\" % (scores[1]*100))\n",
    "\n",
    "# Test and train accuracy of the model\n",
    "model_1_test = scores[1]\n",
    "model_1_train = max(history1.history['acc'])\n",
    "\n",
    "# Plotting Train and Test Loss VS no. of epochs\n",
    "# list of epoch numbers\n",
    "x = list(range(1,31))\n",
    "\n",
    "# Validation loss\n",
    "vy = history1.history['val_loss']\n",
    "# Training loss\n",
    "ty = history1.history['loss']\n",
    "\n",
    "# Calling the function to draw the plot\n",
    "plt_dynamic(x, vy, ty)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Model having 1 LSTM layer with 42 LSTM Units "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_2 (LSTM)                (None, 42)                8736      \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 42)                0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 6)                 258       \n",
      "=================================================================\n",
      "Total params: 8,994\n",
      "Trainable params: 8,994\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "Train on 7352 samples, validate on 2947 samples\n",
      "Epoch 1/30\n",
      "7352/7352 [==============================] - 47s 6ms/step - loss: 1.2947 - acc: 0.4499 - val_loss: 1.1957 - val_acc: 0.4571\n",
      "Epoch 2/30\n",
      "7352/7352 [==============================] - 43s 6ms/step - loss: 1.0003 - acc: 0.5594 - val_loss: 0.8970 - val_acc: 0.6196\n",
      "Epoch 3/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.7672 - acc: 0.6571 - val_loss: 0.7434 - val_acc: 0.6634\n",
      "Epoch 4/30\n",
      "7352/7352 [==============================] - 44s 6ms/step - loss: 0.7791 - acc: 0.6485 - val_loss: 0.7595 - val_acc: 0.6970\n",
      "Epoch 5/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.6105 - acc: 0.7588 - val_loss: 0.6652 - val_acc: 0.7387\n",
      "Epoch 6/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.5855 - acc: 0.7833 - val_loss: 0.5378 - val_acc: 0.7781\n",
      "Epoch 7/30\n",
      "7352/7352 [==============================] - 42s 6ms/step - loss: 0.6056 - acc: 0.8017 - val_loss: 0.5069 - val_acc: 0.8633\n",
      "Epoch 8/30\n",
      "7352/7352 [==============================] - 46s 6ms/step - loss: 0.3929 - acc: 0.8848 - val_loss: 0.5762 - val_acc: 0.8300\n",
      "Epoch 9/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.3592 - acc: 0.8962 - val_loss: 0.4852 - val_acc: 0.8507\n",
      "Epoch 10/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.2607 - acc: 0.9219 - val_loss: 0.3340 - val_acc: 0.8979\n",
      "Epoch 11/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.2327 - acc: 0.9280 - val_loss: 0.3171 - val_acc: 0.8955\n",
      "Epoch 12/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.2309 - acc: 0.9263 - val_loss: 0.8502 - val_acc: 0.8246\n",
      "Epoch 13/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.2203 - acc: 0.9282 - val_loss: 0.4000 - val_acc: 0.8935\n",
      "Epoch 14/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.2015 - acc: 0.9361 - val_loss: 0.3798 - val_acc: 0.8880\n",
      "Epoch 15/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.1964 - acc: 0.9399 - val_loss: 0.3339 - val_acc: 0.9104\n",
      "Epoch 16/30\n",
      "7352/7352 [==============================] - 40s 6ms/step - loss: 0.2422 - acc: 0.9295 - val_loss: 0.3901 - val_acc: 0.8982\n",
      "Epoch 17/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.1763 - acc: 0.9397 - val_loss: 0.2982 - val_acc: 0.9118\n",
      "Epoch 18/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.1564 - acc: 0.9425 - val_loss: 0.4010 - val_acc: 0.9016\n",
      "Epoch 19/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.1671 - acc: 0.9442 - val_loss: 0.3645 - val_acc: 0.9019\n",
      "Epoch 20/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.1623 - acc: 0.9421 - val_loss: 0.3848 - val_acc: 0.9114\n",
      "Epoch 21/30\n",
      "7352/7352 [==============================] - 42s 6ms/step - loss: 0.1604 - acc: 0.9460 - val_loss: 0.4165 - val_acc: 0.9121\n",
      "Epoch 22/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.1496 - acc: 0.9464 - val_loss: 0.3730 - val_acc: 0.9158\n",
      "Epoch 23/30\n",
      "7352/7352 [==============================] - 40s 5ms/step - loss: 0.1710 - acc: 0.9470 - val_loss: 0.3039 - val_acc: 0.9125\n",
      "Epoch 24/30\n",
      "7352/7352 [==============================] - 43s 6ms/step - loss: 0.1478 - acc: 0.9463 - val_loss: 0.3399 - val_acc: 0.9077\n",
      "Epoch 25/30\n",
      "7352/7352 [==============================] - 49s 7ms/step - loss: 0.1519 - acc: 0.9467 - val_loss: 0.4050 - val_acc: 0.9077\n",
      "Epoch 26/30\n",
      "7352/7352 [==============================] - 47s 6ms/step - loss: 0.1525 - acc: 0.9468 - val_loss: 0.2773 - val_acc: 0.9145\n",
      "Epoch 27/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.1489 - acc: 0.9513 - val_loss: 0.4474 - val_acc: 0.9026\n",
      "Epoch 28/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.1361 - acc: 0.9489 - val_loss: 0.2907 - val_acc: 0.9203\n",
      "Epoch 29/30\n",
      "7352/7352 [==============================] - 41s 6ms/step - loss: 0.1542 - acc: 0.9510 - val_loss: 0.4495 - val_acc: 0.9084\n",
      "Epoch 30/30\n",
      "7352/7352 [==============================] - 40s 6ms/step - loss: 0.1580 - acc: 0.9476 - val_loss: 0.4272 - val_acc: 0.9158\n"
     ]
    }
   ],
   "source": [
    "# Initiliazing the sequential model\n",
    "model2 = Sequential()\n",
    "# Configuring the parameters\n",
    "model2.add(LSTM(42, input_shape=(timesteps, input_dim)))\n",
    "# Adding a dropout layer\n",
    "model2.add(Dropout(0.5))\n",
    "# Adding a dense output layer with sigmoid activation\n",
    "model2.add(Dense(n_classes, activation='sigmoid'))\n",
    "print(model2.summary())\n",
    "\n",
    "# Compiling the model\n",
    "model2.compile(loss='categorical_crossentropy',optimizer='rmsprop',metrics=['accuracy'])\n",
    "\n",
    "# Training the model\n",
    "history2 = model2.fit(X_train,Y_train,batch_size=batch_size,validation_data=(X_test, Y_test),epochs=epochs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Score: 0.427173\n",
      "Test Accuracy: 91.584662%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Final evaluation of the model\n",
    "scores2 = model2.evaluate(X_test, Y_test, verbose=0)\n",
    "print(\"Test Score: %f\" % (scores2[0]))\n",
    "print(\"Test Accuracy: %f%%\" % (scores2[1]*100))\n",
    "\n",
    "# Confusion Matrix\n",
    "Y_true = pd.Series([ACTIVITIES[y] for y in np.argmax(Y_test, axis=1)])\n",
    "Y_predictions = pd.Series([ACTIVITIES[y] for y in np.argmax(model2.predict(X_test), axis=1)])\n",
    "\n",
    "# Code for drawing seaborn heatmaps\n",
    "class_names = ['laying','sitting','standing','walking','walking_downstairs','walking_upstairs']\n",
    "df_heatmap = pd.DataFrame(confusion_matrix(Y_true, Y_predictions), index=class_names, columns=class_names )\n",
    "fig = plt.figure(figsize=(10,7))\n",
    "heatmap = sns.heatmap(df_heatmap, annot=True, fmt=\"d\")\n",
    "\n",
    "# Setting tick labels for heatmap\n",
    "heatmap.yaxis.set_ticklabels(heatmap.yaxis.get_ticklabels(), rotation=0, ha='right', fontsize=14)\n",
    "heatmap.xaxis.set_ticklabels(heatmap.xaxis.get_ticklabels(), rotation=90, ha='right', fontsize=14)\n",
    "plt.ylabel('True label',size=18)\n",
    "plt.xlabel('Predicted label',size=18)\n",
    "plt.title(\"Confusion Matrix\\n\",size=24)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Test and train accuracy of the model\n",
    "model_2_test = scores2[1]\n",
    "model_2_train = max(history2.history['acc'])\n",
    "\n",
    "# Plotting Train and Test Loss VS no. of epochs\n",
    "# list of epoch numbers\n",
    "x = list(range(1,31))\n",
    "\n",
    "# Validation loss\n",
    "vy = history2.history['val_loss']\n",
    "# Training loss\n",
    "ty = history2.history['loss']\n",
    "\n",
    "# Calling the function to draw the plot\n",
    "plt_dynamic(x, vy, ty)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Model having 1 LSTM layer with 64 LSTM Units with larger Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_3 (LSTM)                (None, 64)                18944     \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 64)                0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 6)                 390       \n",
      "=================================================================\n",
      "Total params: 19,334\n",
      "Trainable params: 19,334\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "Train on 7352 samples, validate on 2947 samples\n",
      "Epoch 1/30\n",
      "7352/7352 [==============================] - 49s 7ms/step - loss: 1.3289 - acc: 0.4153 - val_loss: 1.1952 - val_acc: 0.4842\n",
      "Epoch 2/30\n",
      "7352/7352 [==============================] - 46s 6ms/step - loss: 1.0808 - acc: 0.5403 - val_loss: 0.9875 - val_acc: 0.5840loss: 1.0830 - acc\n",
      "Epoch 3/30\n",
      "7352/7352 [==============================] - 46s 6ms/step - loss: 0.7958 - acc: 0.6654 - val_loss: 0.8333 - val_acc: 0.6464\n",
      "Epoch 4/30\n",
      "7352/7352 [==============================] - 46s 6ms/step - loss: 0.6894 - acc: 0.7082 - val_loss: 0.7353 - val_acc: 0.6983\n",
      "Epoch 5/30\n",
      "7352/7352 [==============================] - 46s 6ms/step - loss: 0.6138 - acc: 0.7357 - val_loss: 0.8443 - val_acc: 0.6882\n",
      "Epoch 6/30\n",
      "7352/7352 [==============================] - 47s 6ms/step - loss: 0.5816 - acc: 0.7579 - val_loss: 0.6961 - val_acc: 0.7577\n",
      "Epoch 7/30\n",
      "7352/7352 [==============================] - 47s 6ms/step - loss: 0.5487 - acc: 0.7843 - val_loss: 1.4815 - val_acc: 0.6057\n",
      "Epoch 8/30\n",
      "7352/7352 [==============================] - 47s 6ms/step - loss: 0.5442 - acc: 0.8078 - val_loss: 0.7764 - val_acc: 0.7550\n",
      "Epoch 9/30\n",
      "7352/7352 [==============================] - 46s 6ms/step - loss: 0.4581 - acc: 0.8562 - val_loss: 0.5762 - val_acc: 0.8517\n",
      "Epoch 10/30\n",
      "7352/7352 [==============================] - 44s 6ms/step - loss: 0.4103 - acc: 0.8826 - val_loss: 0.5640 - val_acc: 0.8436\n",
      "Epoch 11/30\n",
      "7352/7352 [==============================] - 44s 6ms/step - loss: 0.3345 - acc: 0.9044 - val_loss: 0.4452 - val_acc: 0.8819\n",
      "Epoch 12/30\n",
      "7352/7352 [==============================] - 45s 6ms/step - loss: 0.3613 - acc: 0.9025 - val_loss: 0.3402 - val_acc: 0.9009\n",
      "Epoch 13/30\n",
      "7352/7352 [==============================] - 44s 6ms/step - loss: 0.2813 - acc: 0.9102 - val_loss: 0.3724 - val_acc: 0.9013\n",
      "Epoch 14/30\n",
      "7352/7352 [==============================] - 45s 6ms/step - loss: 0.2385 - acc: 0.9305 - val_loss: 0.3165 - val_acc: 0.9080\n",
      "Epoch 15/30\n",
      "7352/7352 [==============================] - 47s 6ms/step - loss: 0.2228 - acc: 0.9340 - val_loss: 0.3286 - val_acc: 0.9145\n",
      "Epoch 16/30\n",
      "7352/7352 [==============================] - 46s 6ms/step - loss: 0.2520 - acc: 0.9272 - val_loss: 0.8850 - val_acc: 0.8402\n",
      "Epoch 17/30\n",
      "7352/7352 [==============================] - 50s 7ms/step - loss: 0.2275 - acc: 0.9338 - val_loss: 0.3064 - val_acc: 0.9114\n",
      "Epoch 18/30\n",
      "7352/7352 [==============================] - 51s 7ms/step - loss: 0.1871 - acc: 0.9391 - val_loss: 0.3734 - val_acc: 0.9135\n",
      "Epoch 19/30\n",
      "7352/7352 [==============================] - 49s 7ms/step - loss: 0.2178 - acc: 0.9353 - val_loss: 0.3587 - val_acc: 0.9101\n",
      "Epoch 20/30\n",
      "7352/7352 [==============================] - 48s 6ms/step - loss: 0.1981 - acc: 0.9410 - val_loss: 0.3054 - val_acc: 0.9084\n",
      "Epoch 21/30\n",
      "7352/7352 [==============================] - 50s 7ms/step - loss: 0.1817 - acc: 0.9444 - val_loss: 0.3525 - val_acc: 0.9030\n",
      "Epoch 22/30\n",
      "7352/7352 [==============================] - 45s 6ms/step - loss: 0.2005 - acc: 0.9410 - val_loss: 0.2881 - val_acc: 0.9264\n",
      "Epoch 23/30\n",
      "7352/7352 [==============================] - 46s 6ms/step - loss: 0.1742 - acc: 0.9431 - val_loss: 0.4975 - val_acc: 0.9043\n",
      "Epoch 24/30\n",
      "7352/7352 [==============================] - 45s 6ms/step - loss: 0.1950 - acc: 0.9422 - val_loss: 0.3071 - val_acc: 0.9189\n",
      "Epoch 25/30\n",
      "7352/7352 [==============================] - 44s 6ms/step - loss: 0.1665 - acc: 0.9457 - val_loss: 0.4143 - val_acc: 0.8904\n",
      "Epoch 26/30\n",
      "7352/7352 [==============================] - 44s 6ms/step - loss: 0.1777 - acc: 0.9453 - val_loss: 0.7865 - val_acc: 0.8761\n",
      "Epoch 27/30\n",
      "7352/7352 [==============================] - 44s 6ms/step - loss: 0.1783 - acc: 0.9440 - val_loss: 0.3182 - val_acc: 0.9148\n",
      "Epoch 28/30\n",
      "7352/7352 [==============================] - 45s 6ms/step - loss: 0.1646 - acc: 0.9483 - val_loss: 0.2716 - val_acc: 0.9247\n",
      "Epoch 29/30\n",
      "7352/7352 [==============================] - 44s 6ms/step - loss: 0.2131 - acc: 0.9422 - val_loss: 0.3630 - val_acc: 0.9046\n",
      "Epoch 30/30\n",
      "7352/7352 [==============================] - 44s 6ms/step - loss: 0.2133 - acc: 0.9359 - val_loss: 0.2799 - val_acc: 0.9148\n"
     ]
    }
   ],
   "source": [
    "# Initiliazing the sequential model\n",
    "model3 = Sequential()\n",
    "# Configuring the parameters\n",
    "model3.add(LSTM(64, input_shape=(timesteps, input_dim)))\n",
    "# Adding a dropout layer\n",
    "model3.add(Dropout(0.7))\n",
    "# Adding a dense output layer with sigmoid activation\n",
    "model3.add(Dense(n_classes, activation='sigmoid'))\n",
    "print(model3.summary())\n",
    "\n",
    "# Compiling the model\n",
    "model3.compile(loss='categorical_crossentropy',optimizer='rmsprop',metrics=['accuracy'])\n",
    "\n",
    "# Training the model\n",
    "history3 = model3.fit(X_train,Y_train,batch_size=batch_size,validation_data=(X_test, Y_test),epochs=epochs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Score: 0.279943\n",
      "Test Accuracy: 91.482864%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Final evaluation of the model\n",
    "scores3 = model3.evaluate(X_test, Y_test, verbose=0)\n",
    "print(\"Test Score: %f\" % (scores3[0]))\n",
    "print(\"Test Accuracy: %f%%\" % (scores3[1]*100))\n",
    "\n",
    "# Confusion Matrix\n",
    "Y_true = pd.Series([ACTIVITIES[y] for y in np.argmax(Y_test, axis=1)])\n",
    "Y_predictions = pd.Series([ACTIVITIES[y] for y in np.argmax(model3.predict(X_test), axis=1)])\n",
    "\n",
    "# Code for drawing seaborn heatmaps\n",
    "class_names = ['laying','sitting','standing','walking','walking_downstairs','walking_upstairs']\n",
    "df_heatmap = pd.DataFrame(confusion_matrix(Y_true, Y_predictions), index=class_names, columns=class_names )\n",
    "fig = plt.figure(figsize=(10,7))\n",
    "heatmap = sns.heatmap(df_heatmap, annot=True, fmt=\"d\")\n",
    "\n",
    "# Setting tick labels for heatmap\n",
    "heatmap.yaxis.set_ticklabels(heatmap.yaxis.get_ticklabels(), rotation=0, ha='right', fontsize=14)\n",
    "heatmap.xaxis.set_ticklabels(heatmap.xaxis.get_ticklabels(), rotation=90, ha='right', fontsize=14)\n",
    "plt.ylabel('True label',size=18)\n",
    "plt.xlabel('Predicted label',size=18)\n",
    "plt.title(\"Confusion Matrix\\n\",size=24)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Test and train accuracy of the model\n",
    "model_3_test = scores3[1]\n",
    "model_3_train = max(history3.history['acc'])\n",
    "\n",
    "# Plotting Train and Test Loss VS no. of epochs\n",
    "# list of epoch numbers\n",
    "x = list(range(1,31))\n",
    "\n",
    "# Validation loss\n",
    "vy = history3.history['val_loss']\n",
    "# Training loss\n",
    "ty = history3.history['loss']\n",
    "\n",
    "# Calling the function to draw the plot\n",
    "plt_dynamic(x, vy, ty)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Model having 2 LSTM layer with 32 LSTM Units "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_4 (LSTM)                (None, 128, 32)           5376      \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 128, 32)           0         \n",
      "_________________________________________________________________\n",
      "lstm_5 (LSTM)                (None, 32)                8320      \n",
      "_________________________________________________________________\n",
      "dropout_5 (Dropout)          (None, 32)                0         \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 6)                 198       \n",
      "=================================================================\n",
      "Total params: 13,894\n",
      "Trainable params: 13,894\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "Train on 7352 samples, validate on 2947 samples\n",
      "Epoch 1/30\n",
      "7352/7352 [==============================] - 88s 12ms/step - loss: 1.2283 - acc: 0.4827 - val_loss: 1.0618 - val_acc: 0.4866\n",
      "Epoch 2/30\n",
      "7352/7352 [==============================] - 83s 11ms/step - loss: 0.8534 - acc: 0.6099 - val_loss: 1.0135 - val_acc: 0.5456\n",
      "Epoch 3/30\n",
      "7352/7352 [==============================] - 81s 11ms/step - loss: 0.7211 - acc: 0.6493 - val_loss: 0.7843 - val_acc: 0.6094\n",
      "Epoch 4/30\n",
      "7352/7352 [==============================] - 82s 11ms/step - loss: 0.6762 - acc: 0.6669 - val_loss: 0.7229 - val_acc: 0.6179\n",
      "Epoch 5/30\n",
      "7352/7352 [==============================] - 81s 11ms/step - loss: 0.6600 - acc: 0.6647 - val_loss: 0.6788 - val_acc: 0.6210\n",
      "Epoch 6/30\n",
      "7352/7352 [==============================] - 82s 11ms/step - loss: 0.5906 - acc: 0.7160 - val_loss: 0.6746 - val_acc: 0.7296\n",
      "Epoch 7/30\n",
      "7352/7352 [==============================] - 82s 11ms/step - loss: 0.5347 - acc: 0.7671 - val_loss: 0.5615 - val_acc: 0.7370\n",
      "Epoch 8/30\n",
      "7352/7352 [==============================] - 80s 11ms/step - loss: 0.4404 - acc: 0.7961 - val_loss: 0.5352 - val_acc: 0.7384\n",
      "Epoch 9/30\n",
      "7352/7352 [==============================] - 80s 11ms/step - loss: 0.3666 - acc: 0.8550 - val_loss: 0.5541 - val_acc: 0.8405\n",
      "Epoch 10/30\n",
      "7352/7352 [==============================] - 81s 11ms/step - loss: 0.3070 - acc: 0.9064 - val_loss: 0.6806 - val_acc: 0.8079\n",
      "Epoch 11/30\n",
      "7352/7352 [==============================] - 85s 12ms/step - loss: 0.2319 - acc: 0.9298 - val_loss: 0.4894 - val_acc: 0.8748\n",
      "Epoch 12/30\n",
      "7352/7352 [==============================] - 84s 11ms/step - loss: 0.2052 - acc: 0.9350 - val_loss: 0.5480 - val_acc: 0.8755\n",
      "Epoch 13/30\n",
      "7352/7352 [==============================] - 83s 11ms/step - loss: 0.1892 - acc: 0.9414 - val_loss: 0.4811 - val_acc: 0.8761\n",
      "Epoch 14/30\n",
      "7352/7352 [==============================] - 80s 11ms/step - loss: 0.1807 - acc: 0.9387 - val_loss: 0.4882 - val_acc: 0.8728\n",
      "Epoch 15/30\n",
      "7352/7352 [==============================] - 79s 11ms/step - loss: 0.1653 - acc: 0.9460 - val_loss: 0.4981 - val_acc: 0.8799\n",
      "Epoch 16/30\n",
      "7352/7352 [==============================] - 79s 11ms/step - loss: 0.1518 - acc: 0.9457 - val_loss: 0.4923 - val_acc: 0.8941\n",
      "Epoch 17/30\n",
      "7352/7352 [==============================] - 80s 11ms/step - loss: 0.1593 - acc: 0.9479 - val_loss: 0.5772 - val_acc: 0.8884\n",
      "Epoch 18/30\n",
      "7352/7352 [==============================] - 82s 11ms/step - loss: 0.1574 - acc: 0.9436 - val_loss: 0.4696 - val_acc: 0.8924\n",
      "Epoch 19/30\n",
      "7352/7352 [==============================] - 80s 11ms/step - loss: 0.1415 - acc: 0.9502 - val_loss: 0.4920 - val_acc: 0.8938\n",
      "Epoch 20/30\n",
      "7352/7352 [==============================] - 83s 11ms/step - loss: 0.1528 - acc: 0.9502 - val_loss: 0.5691 - val_acc: 0.8928\n",
      "Epoch 21/30\n",
      "7352/7352 [==============================] - 83s 11ms/step - loss: 0.1461 - acc: 0.9494 - val_loss: 0.6179 - val_acc: 0.8856\n",
      "Epoch 22/30\n",
      "7352/7352 [==============================] - 80s 11ms/step - loss: 0.1449 - acc: 0.9531 - val_loss: 0.4345 - val_acc: 0.9019\n",
      "Epoch 23/30\n",
      "7352/7352 [==============================] - 82s 11ms/step - loss: 0.1583 - acc: 0.9486 - val_loss: 0.5218 - val_acc: 0.8870\n",
      "Epoch 24/30\n",
      "7352/7352 [==============================] - 81s 11ms/step - loss: 0.1414 - acc: 0.9494 - val_loss: 0.4950 - val_acc: 0.8985\n",
      "Epoch 25/30\n",
      "7352/7352 [==============================] - 81s 11ms/step - loss: 0.1645 - acc: 0.9480 - val_loss: 0.4609 - val_acc: 0.9002\n",
      "Epoch 26/30\n",
      "7352/7352 [==============================] - 81s 11ms/step - loss: 0.1409 - acc: 0.9480 - val_loss: 0.5459 - val_acc: 0.8867\n",
      "Epoch 27/30\n",
      "7352/7352 [==============================] - 82s 11ms/step - loss: 0.1572 - acc: 0.9461 - val_loss: 0.5245 - val_acc: 0.8907\n",
      "Epoch 28/30\n",
      "7352/7352 [==============================] - 86s 12ms/step - loss: 0.1379 - acc: 0.9527 - val_loss: 0.5630 - val_acc: 0.8996\n",
      "Epoch 29/30\n",
      "7352/7352 [==============================] - 86s 12ms/step - loss: 0.1426 - acc: 0.9529 - val_loss: 0.6665 - val_acc: 0.8887\n",
      "Epoch 30/30\n",
      "7352/7352 [==============================] - 88s 12ms/step - loss: 0.1608 - acc: 0.9509 - val_loss: 0.5688 - val_acc: 0.9067\n"
     ]
    }
   ],
   "source": [
    "# Initiliazing the sequential model\n",
    "model4 = Sequential()\n",
    "# Configuring the parameters\n",
    "model4.add(LSTM(32,return_sequences=True, input_shape=(timesteps, input_dim)))\n",
    "# Adding a dropout layer\n",
    "model4.add(Dropout(0.5))\n",
    "\n",
    "# Configuring the parameters\n",
    "model4.add(LSTM(32))\n",
    "# Adding a dropout layer\n",
    "model4.add(Dropout(0.5))\n",
    "# Adding a dense output layer with sigmoid activation\n",
    "model4.add(Dense(n_classes, activation='sigmoid'))\n",
    "print(model4.summary())\n",
    "\n",
    "# Compiling the model\n",
    "model4.compile(loss='categorical_crossentropy',optimizer='rmsprop',metrics=['accuracy'])\n",
    "\n",
    "# Training the model\n",
    "history4 = model4.fit(X_train,Y_train,batch_size=batch_size,validation_data=(X_test, Y_test),epochs=epochs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Score: 0.568783\n",
      "Test Accuracy: 90.668476%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Final evaluation of the model\n",
    "scores4 = model4.evaluate(X_test, Y_test, verbose=0)\n",
    "print(\"Test Score: %f\" % (scores4[0]))\n",
    "print(\"Test Accuracy: %f%%\" % (scores4[1]*100))\n",
    "\n",
    "# Confusion Matrix\n",
    "Y_true = pd.Series([ACTIVITIES[y] for y in np.argmax(Y_test, axis=1)])\n",
    "Y_predictions = pd.Series([ACTIVITIES[y] for y in np.argmax(model4.predict(X_test), axis=1)])\n",
    "\n",
    "# Code for drawing seaborn heatmaps\n",
    "class_names = ['laying','sitting','standing','walking','walking_downstairs','walking_upstairs']\n",
    "df_heatmap = pd.DataFrame(confusion_matrix(Y_true, Y_predictions), index=class_names, columns=class_names )\n",
    "fig = plt.figure(figsize=(10,7))\n",
    "heatmap = sns.heatmap(df_heatmap, annot=True, fmt=\"d\")\n",
    "\n",
    "# Setting tick labels for heatmap\n",
    "heatmap.yaxis.set_ticklabels(heatmap.yaxis.get_ticklabels(), rotation=0, ha='right', fontsize=14)\n",
    "heatmap.xaxis.set_ticklabels(heatmap.xaxis.get_ticklabels(), rotation=90, ha='right', fontsize=14)\n",
    "plt.ylabel('True label',size=18)\n",
    "plt.xlabel('Predicted label',size=18)\n",
    "plt.title(\"Confusion Matrix\\n\",size=24)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Test and train accuracy of the model\n",
    "model_4_test = scores4[1]\n",
    "model_4_train = max(history4.history['acc'])\n",
    "\n",
    "# Plotting Train and Test Loss VS no. of epochs\n",
    "# list of epoch numbers\n",
    "x = list(range(1,31))\n",
    "\n",
    "# Validation loss\n",
    "vy = history4.history['val_loss']\n",
    "# Training loss\n",
    "ty = history4.history['loss']\n",
    "\n",
    "# Calling the function to draw the plot\n",
    "plt_dynamic(x, vy, ty)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Model having 2 LSTM layer with 64 LSTM Units and with larger Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_6 (LSTM)                (None, 128, 64)           18944     \n",
      "_________________________________________________________________\n",
      "dropout_6 (Dropout)          (None, 128, 64)           0         \n",
      "_________________________________________________________________\n",
      "lstm_7 (LSTM)                (None, 64)                33024     \n",
      "_________________________________________________________________\n",
      "dropout_7 (Dropout)          (None, 64)                0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 6)                 390       \n",
      "=================================================================\n",
      "Total params: 52,358\n",
      "Trainable params: 52,358\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n",
      "Train on 7352 samples, validate on 2947 samples\n",
      "Epoch 1/30\n",
      "7352/7352 [==============================] - 113s 15ms/step - loss: 1.2066 - acc: 0.4969 - val_loss: 1.0893 - val_acc: 0.5409\n",
      "Epoch 2/30\n",
      "7352/7352 [==============================] - 117s 16ms/step - loss: 0.7975 - acc: 0.6779 - val_loss: 1.0566 - val_acc: 0.5908\n",
      "Epoch 3/30\n",
      "7352/7352 [==============================] - 110s 15ms/step - loss: 0.5726 - acc: 0.7757 - val_loss: 0.5347 - val_acc: 0.8320\n",
      "Epoch 4/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.3818 - acc: 0.8853 - val_loss: 0.4258 - val_acc: 0.8707\n",
      "Epoch 5/30\n",
      "7352/7352 [==============================] - 109s 15ms/step - loss: 0.3508 - acc: 0.9032 - val_loss: 0.6369 - val_acc: 0.8639\n",
      "Epoch 6/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.2589 - acc: 0.9233 - val_loss: 0.6050 - val_acc: 0.8748\n",
      "Epoch 7/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.2256 - acc: 0.9276 - val_loss: 0.3134 - val_acc: 0.8975\n",
      "Epoch 8/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.2193 - acc: 0.9334 - val_loss: 0.7002 - val_acc: 0.8802\n",
      "Epoch 9/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.2012 - acc: 0.9313 - val_loss: 0.4142 - val_acc: 0.9067\n",
      "Epoch 10/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.1901 - acc: 0.9406 - val_loss: 0.4822 - val_acc: 0.8907\n",
      "Epoch 11/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.1934 - acc: 0.9373 - val_loss: 0.3472 - val_acc: 0.9162\n",
      "Epoch 12/30\n",
      "7352/7352 [==============================] - 111s 15ms/step - loss: 0.1843 - acc: 0.9414 - val_loss: 0.3798 - val_acc: 0.9125\n",
      "Epoch 13/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.1649 - acc: 0.9433 - val_loss: 0.4329 - val_acc: 0.9148\n",
      "Epoch 14/30\n",
      "7352/7352 [==============================] - 110s 15ms/step - loss: 0.1746 - acc: 0.9445 - val_loss: 0.5866 - val_acc: 0.8979\n",
      "Epoch 15/30\n",
      "7352/7352 [==============================] - 111s 15ms/step - loss: 0.2108 - acc: 0.9400 - val_loss: 0.5632 - val_acc: 0.9053\n",
      "Epoch 16/30\n",
      "7352/7352 [==============================] - 112s 15ms/step - loss: 0.1627 - acc: 0.9453 - val_loss: 0.4469 - val_acc: 0.9087\n",
      "Epoch 17/30\n",
      "7352/7352 [==============================] - 110s 15ms/step - loss: 0.1677 - acc: 0.9441 - val_loss: 0.6730 - val_acc: 0.9016\n",
      "Epoch 18/30\n",
      "7352/7352 [==============================] - 110s 15ms/step - loss: 0.1597 - acc: 0.9464 - val_loss: 0.6454 - val_acc: 0.9040\n",
      "Epoch 19/30\n",
      "7352/7352 [==============================] - 113s 15ms/step - loss: 0.1621 - acc: 0.9448 - val_loss: 0.4840 - val_acc: 0.9141\n",
      "Epoch 20/30\n",
      "7352/7352 [==============================] - 110s 15ms/step - loss: 0.1776 - acc: 0.9461 - val_loss: 0.6172 - val_acc: 0.9023\n",
      "Epoch 21/30\n",
      "7352/7352 [==============================] - 112s 15ms/step - loss: 0.1664 - acc: 0.9437 - val_loss: 0.4075 - val_acc: 0.9141\n",
      "Epoch 22/30\n",
      "7352/7352 [==============================] - 110s 15ms/step - loss: 0.1388 - acc: 0.9474 - val_loss: 0.5453 - val_acc: 0.9182\n",
      "Epoch 23/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.1613 - acc: 0.9489 - val_loss: 0.6258 - val_acc: 0.8921\n",
      "Epoch 24/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.1938 - acc: 0.9400 - val_loss: 0.4679 - val_acc: 0.9145\n",
      "Epoch 25/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.1543 - acc: 0.9463 - val_loss: 0.4444 - val_acc: 0.9175\n",
      "Epoch 26/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.1556 - acc: 0.9460 - val_loss: 0.4261 - val_acc: 0.9189\n",
      "Epoch 27/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.1468 - acc: 0.9471 - val_loss: 0.4721 - val_acc: 0.9135\n",
      "Epoch 28/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.1600 - acc: 0.9456 - val_loss: 0.6335 - val_acc: 0.8968\n",
      "Epoch 29/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.2024 - acc: 0.9434 - val_loss: 0.5230 - val_acc: 0.8965\n",
      "Epoch 30/30\n",
      "7352/7352 [==============================] - 108s 15ms/step - loss: 0.2431 - acc: 0.9380 - val_loss: 0.5654 - val_acc: 0.8968\n"
     ]
    }
   ],
   "source": [
    "# Initiliazing the sequential model\n",
    "model5 = Sequential()\n",
    "# Configuring the parameters\n",
    "model5.add(LSTM(64,return_sequences=True, input_shape=(timesteps, input_dim)))\n",
    "# Adding a dropout layer\n",
    "model5.add(Dropout(0.7))\n",
    "\n",
    "# Configuring the parameters\n",
    "model5.add(LSTM(64))\n",
    "# Adding a dropout layer\n",
    "model5.add(Dropout(0.7))\n",
    "# Adding a dense output layer with sigmoid activation\n",
    "model5.add(Dense(n_classes, activation='sigmoid'))\n",
    "print(model5.summary())\n",
    "\n",
    "# Compiling the model\n",
    "model5.compile(loss='categorical_crossentropy',optimizer='rmsprop',metrics=['accuracy'])\n",
    "\n",
    "# Training the model\n",
    "history5 = model5.fit(X_train,Y_train,batch_size=batch_size,validation_data=(X_test, Y_test),epochs=epochs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Score: 0.565438\n",
      "Test Accuracy: 89.684425%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Final evaluation of the model\n",
    "scores5 = model5.evaluate(X_test, Y_test, verbose=0)\n",
    "print(\"Test Score: %f\" % (scores5[0]))\n",
    "print(\"Test Accuracy: %f%%\" % (scores5[1]*100))\n",
    "\n",
    "# Confusion Matrix\n",
    "Y_true = pd.Series([ACTIVITIES[y] for y in np.argmax(Y_test, axis=1)])\n",
    "Y_predictions = pd.Series([ACTIVITIES[y] for y in np.argmax(model5.predict(X_test), axis=1)])\n",
    "\n",
    "# Code for drawing seaborn heatmaps\n",
    "class_names = ['laying','sitting','standing','walking','walking_downstairs','walking_upstairs']\n",
    "df_heatmap = pd.DataFrame(confusion_matrix(Y_true, Y_predictions), index=class_names, columns=class_names )\n",
    "fig = plt.figure(figsize=(10,7))\n",
    "heatmap = sns.heatmap(df_heatmap, annot=True, fmt=\"d\")\n",
    "\n",
    "# Setting tick labels for heatmap\n",
    "heatmap.yaxis.set_ticklabels(heatmap.yaxis.get_ticklabels(), rotation=0, ha='right', fontsize=14)\n",
    "heatmap.xaxis.set_ticklabels(heatmap.xaxis.get_ticklabels(), rotation=90, ha='right', fontsize=14)\n",
    "plt.ylabel('True label',size=18)\n",
    "plt.xlabel('Predicted label',size=18)\n",
    "plt.title(\"Confusion Matrix\\n\",size=24)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Test and train accuracy of the model\n",
    "model_5_test = scores5[1]\n",
    "model_5_train = max(history5.history['acc'])\n",
    "\n",
    "# Plotting Train and Test Loss VS no. of epochs\n",
    "# list of epoch numbers\n",
    "x = list(range(1,31))\n",
    "\n",
    "# Validation loss\n",
    "vy = history5.history['val_loss']\n",
    "# Training loss\n",
    "ty = history5.history['loss']\n",
    "\n",
    "# Calling the function to draw the plot\n",
    "plt_dynamic(x, vy, ty)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performance Table\n",
      "+----------------------------------------------------------------------+--------------------+--------------------+\n",
      "|                                Models                                |       Train        |        Test        |\n",
      "+----------------------------------------------------------------------+--------------------+--------------------+\n",
      "|             Model having 1 LSTM layer with 32 LSTM Units             | 0.9416485310119695 | 0.8829317950458093 |\n",
      "|             Model having 1 LSTM layer with 42 LSTM Units             | 0.9513057671381937 | 0.9158466236851035 |\n",
      "|   Model having 1 LSTM layer with 64 LSTM Units with larger Dropout   | 0.9483133841131665 | 0.9148286392941974 |\n",
      "|             Model having 2 LSTM layer with 32 LSTM Units             | 0.9530739934711643 | 0.9066847641669494 |\n",
      "| Model having 2 LSTM layer with 64 LSTM Units and with larger Dropout | 0.9488574537540805 | 0.8968442483881914 |\n",
      "+----------------------------------------------------------------------+--------------------+--------------------+\n"
     ]
    }
   ],
   "source": [
    "from prettytable import PrettyTable\n",
    "\n",
    "print('Performance Table')\n",
    "x = PrettyTable()\n",
    "x.field_names =[\"Models\",\"Train\",\"Test\"]\n",
    "\n",
    "x.add_row([\"Model having 1 LSTM layer with 32 LSTM Units\",model_1_train,model_1_test])\n",
    "x.add_row([\"Model having 1 LSTM layer with 42 LSTM Units\",model_2_train,model_2_test])\n",
    "x.add_row([\"Model having 1 LSTM layer with 64 LSTM Units with larger Dropout\",model_3_train,model_3_test])\n",
    "x.add_row([\"Model having 2 LSTM layer with 32 LSTM Units\",model_4_train,model_4_test])\n",
    "x.add_row([\"Model having 2 LSTM layer with 64 LSTM Units and with larger Dropout\",model_5_train,model_5_test])\n",
    "\n",
    "print(x)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}