group79_other_algorithms1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "_FKDoCYlj7Wt"
   },
   "source": [
    "\n",
    "\n",
    "---\n",
    "**Model 1**: DecisionTreeRegressor with log transformation and upsampling and also feature engineering.\n",
    "\n",
    "**To check the performance, simply run the following code cell, and result should be print out immediately.**\n",
    "\n",
    "the code for plotting and visualisation are also followed\n",
    "\n",
    "usually the code will load file named forstfires.csv in current directory or under data directory, still not found the code will download the data set from Haichen's Github."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 747
    },
    "colab_type": "code",
    "id": "tYyQuWbaCG75",
    "outputId": "96331126-9a4c-490f-a1ad-941ee1618b03"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "File loaded online! From:  https://raw.githubusercontent.com/haichenzhu/Forest-Fire-Area-Prediction/master/forestfires.csv\n",
      "Start training DecisionTreeRegressor(\n",
      "\trandom_state=123, \n",
      "\tmax_depth=9,      \n",
      "\tmax_features=3,\n",
      "\tmin_samples_leaf=2, \n",
      "\tmin_samples_split=2,      \n",
      "\tmin_weight_fraction_leaf=0)...\n",
      "Training finished! Start predicting...\n",
      "The Performance of Our DecisionTreeRegressor:\n",
      "MSE:   4809.49\n",
      "RMSE:  69.35\n",
      "MAD:   8.297\n",
      "MAE:   18.988\n",
      "NLL:   1341.669\n",
      "RR:    -0.075\n",
      "AUC:   0.593\n",
      "Time cost in seconds:  0.11200690269470215 s.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.feature_selection import RFE, RFECV\n",
    "from sklearn.preprocessing import LabelEncoder, OneHotEncoder, QuantileTransformer\n",
    "from sklearn.model_selection import cross_val_score, KFold, GridSearchCV, StratifiedShuffleSplit, train_test_split\n",
    "from sklearn.tree import DecisionTreeRegressor, export_graphviz\n",
    "from sklearn.externals.six import StringIO\n",
    "from IPython.display import Image\n",
    "from sklearn.metrics import mean_squared_error, log_loss, r2_score, mean_absolute_error\n",
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "from scipy.integrate import simps\n",
    "from sklearn.utils import resample\n",
    "import time\n",
    "%matplotlib inline\n",
    "start_time = time.time()\n",
    "\n",
    "\n",
    "def mean_absolute_deviation(y_predict):\n",
    "    '''\n",
    "    the mean absolue deviation of every prediction to the prediction mean, defined as:\n",
    "    mean(abs(p - y_predict.mean() for p in y_predict))\n",
    "    '''\n",
    "    prediction_mean = np.mean(y_predict)\n",
    "    distance = np.abs(y_predict - prediction_mean)\n",
    "    return np.sum(distance) / len(y_predict)\n",
    "\n",
    "\n",
    "def negative_log_likelihood(y_true, y_predict):\n",
    "    '''\n",
    "    defined as : −log𝑝(𝑦|𝐷,𝑥), \n",
    "    where y is the true value, D is the dataset, x represents a query point\n",
    "    '''\n",
    "    prediction_mean = np.mean(y_predict)\n",
    "    prediction_variance = np.var(y_predict)\n",
    "    return np.log(2 * np.pi * prediction_variance) / 2 + \\\n",
    "            np.sum((y_true - prediction_mean) ** 2) / (2 * prediction_variance)\n",
    "\n",
    "\n",
    "def rec(y_true, y_predict):\n",
    "    '''\n",
    "    Regression Error Characteristic curve with X-axis represents deviation, and y-axis shows \n",
    "    the current accuracy under deviation, return the range of deviation, the accuracy of every\n",
    "    deviation, the area under the curve\n",
    "    '''\n",
    "    accuracy = []\n",
    "    begin = 0\n",
    "    end = 15\n",
    "    interval = 0.1   \n",
    "    epsilon = np.arange(begin, end, interval)\n",
    "    \n",
    "    # loop to find the accuracy for every deviation\n",
    "    for i in range(len(epsilon)):\n",
    "        count = 0\n",
    "        for j in range(len(y_true)):\n",
    "            if np.abs(y_true[j] - y_predict[j]) < epsilon[i]:\n",
    "                count = count + 1\n",
    "        # record current accuracy value \n",
    "        accuracy.append(count/len(y_true))\n",
    "    \n",
    "    # using Simpson's rule to calculating area-under-curve\n",
    "    area_under_curve = simps(accuracy, epsilon) / end\n",
    "   \n",
    "    return epsilon, accuracy, area_under_curve\n",
    "\n",
    "\n",
    "# read data set into a pandas DataFrame\n",
    "try:\n",
    "    data = pd.read_csv('forestfires.csv')\n",
    "    print(\"File loaded locally! \")\n",
    "except FileNotFoundError:\n",
    "    try:\n",
    "        data = pd.read_csv('./data/forestfires.csv')\n",
    "        print(\"File loaded locally! \")\n",
    "    except FileNotFoundError:\n",
    "        url = 'https://raw.githubusercontent.com/haichenzhu/Forest-Fire-Area-Prediction/master/forestfires.csv'\n",
    "        data = pd.read_csv(url)\n",
    "        print(\"File loaded online! From: \", url)\n",
    "except:\n",
    "    print(\"No file named forestfires.csv in the ./data directory or no Internet access! Mission Failed! \")\n",
    "\n",
    "# these months have very small vrecords so remove from the dataset\n",
    "months_to_remove = ['nov', 'may', 'jan']\n",
    "data = data.drop(data[data['month'].isin(months_to_remove)].index, axis=0, inplace=False)\n",
    "\n",
    "# implementing log transformation to the target value\n",
    "data['Logarea'] = np.log(data['area'] + 1)\n",
    "\n",
    "# make bins to classifier the target value for equal train test set split\n",
    "data['Logarea_bins'] = pd.cut(data['Logarea'], bins=[0, 1, 2, 3, 4, 8], include_lowest=True, \n",
    "                                 labels=['0-1', '1-2', '2-3', '3-4', '>4'])\n",
    "\n",
    "# change object type into integer type\n",
    "labelencoder = LabelEncoder()\n",
    "data['month'] = labelencoder.fit_transform(data['month'])\n",
    "data['day'] = labelencoder.fit_transform(data['day'])\n",
    "\n",
    "# using StratifiedShuffleSplit spliter to equally split the data set accroding to Logarea_bins\n",
    "train_index = test_index = data.index\n",
    "sss = StratifiedShuffleSplit(random_state=0, n_splits=1, test_size=0.3)\n",
    "for train_i, test_i in sss.split(data.values, data.Logarea_bins.values):\n",
    "    train_index = train_i\n",
    "    test_index = test_i\n",
    "\n",
    "# upsampling each bin of Logarea\n",
    "data_4 = data[data.index.isin(train_index)][data[data.index.isin(train_index)].Logarea_bins == '>4']\n",
    "data_34 = data[data.index.isin(train_index)][data[data.index.isin(train_index)].Logarea_bins == '3-4']\n",
    "data_23 = data[data.index.isin(train_index)][data[data.index.isin(train_index)].Logarea_bins == '2-3']\n",
    "data_12 = data[data.index.isin(train_index)][data[data.index.isin(train_index)].Logarea_bins == '1-2']\n",
    "\n",
    "# upsampling with different numbers to reduce the imbalance in the dataset\n",
    "data_4_upsampled = resample(data_4, replace=True, n_samples=80, random_state=123)\n",
    "data_34_upsampled = resample(data_34, replace=True, n_samples=120, random_state=123)\n",
    "data_23_upsampled = resample(data_23, replace=True, n_samples=160, random_state=123)\n",
    "data_12_upsampled = resample(data_12, replace=True, n_samples=200, random_state=123)\n",
    "\n",
    "# combine upsampled data in axis=0\n",
    "data_upsampled = pd.concat([data, data_4_upsampled, data_34_upsampled, data_23_upsampled, data_12_upsampled])\n",
    "\n",
    "# drop irrelevant features according to RFE method, code can be found downward\n",
    "features = data_upsampled.drop(['area', 'Logarea', 'Logarea_bins', 'month', 'day', 'FFMC', 'rain'], axis=1)\n",
    "outputs = data_upsampled['Logarea']\n",
    "\n",
    "# make sure every upsampled data are copied using index.isin method\n",
    "X_train = features[features.index.isin(train_index)]\n",
    "y_train = outputs[outputs.index.isin(train_index)]\n",
    "X_test = features[features.index.isin(test_index)]\n",
    "y_test = outputs[outputs.index.isin(test_index)]\n",
    "\n",
    "# have fine tune the parameters using GirdSearchCV() method, code can be found downwards\n",
    "print(\"Start training DecisionTreeRegressor(\\n\\trandom_state=123, \\n\\tmax_depth=9, \\\n",
    "     \\n\\tmax_features=3,\\n\\tmin_samples_leaf=2, \\n\\tmin_samples_split=2, \\\n",
    "     \\n\\tmin_weight_fraction_leaf=0)...\")\n",
    "dtr = DecisionTreeRegressor(random_state=123, \n",
    "                            max_depth=9, \n",
    "                            max_features=3, \n",
    "                            min_samples_leaf=2, \n",
    "                            min_samples_split=2, \n",
    "                            min_weight_fraction_leaf=0)\n",
    "dtr.fit(X_train, y_train)\n",
    "\n",
    "# make prediction\n",
    "print('Training finished! Start predicting...')\n",
    "# do inverse of log transformation\n",
    "y_predict = np.e ** dtr.predict(X_test) - 1\n",
    "y_true = np.e ** y_test - 1\n",
    "\n",
    "# print out the performance of our model\n",
    "print('The Performance of Our DecisionTreeRegressor:')\n",
    "mse = mean_squared_error(y_true, y_predict)\n",
    "rmse = np.sqrt(mse)\n",
    "print('MSE:  ', mse.round(3))\n",
    "print('RMSE: ', rmse.round(3))\n",
    "print('MAD:  ', mean_absolute_deviation(y_predict).round(3))\n",
    "print('MAE:  ', mean_absolute_error(y_true, y_predict).round(3))\n",
    "print('NLL:  ', negative_log_likelihood(y_true, y_predict).round(3))\n",
    "\n",
    "rr = r2_score(y_true, y_predict)\n",
    "print('RR:   ', rr.round(3))\n",
    "\n",
    "deviation, accuracy, area_under_curve = rec(y_true.to_numpy(), y_predict)\n",
    "print('AUC:  ', area_under_curve.round(3))\n",
    "\n",
    "end_time = time.time() - start_time\n",
    "print('Time cost in seconds: ', end_time, 's.')\n",
    "\n",
    "\n",
    "# Plotting\n",
    "plt.figure(figsize=(13 , 6))\n",
    "\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.title(\"Zoom in small forest fires\")\n",
    "plt.scatter(y_true, y_predict, color=\"red\")\n",
    "plt.xlabel(\"True Target\")\n",
    "plt.ylabel(\"Predicted\")\n",
    "plt.xlim(-1, 30)\n",
    "plt.ylim(-1, 30)\n",
    "plt.plot([y_true.min(), y_true.max()], [y_true.min(), y_true.max()], 'k--', lw=4)\n",
    "plt.text(20, 15, r\"$R^2 = %0.4f$\" %rr, fontsize=15)\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.title(\"Regression Error Characteristic (REC)\")\n",
    "plt.plot(deviation, accuracy, \"--y\", lw =3)\n",
    "plt.xlabel(\"Deviation\")\n",
    "plt.ylabel(\"Accuracy (%)\")\n",
    "plt.text(10, 0.09, \"AUC = %0.4f\" %area_under_curve, fontsize=15)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "CZx5R3mjl1pH"
   },
   "source": [
    "\n",
    "                                          code above for testing\n",
    "\n",
    "---\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "1CgH0yFQCG7_",
    "outputId": "b98e9aa8-1b39-4c7a-ac6f-405baf555c45"
   },
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"forestfires.csv\")\n",
    "\n",
    "data.describe()\n",
    "data.corr()\n",
    "print(data.info())\n",
    "data.isnull().sum().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "0jZLLiLdCG8D",
    "outputId": "6d7956f0-ff71-4b69-841f-2980568137c2"
   },
   "outputs": [],
   "source": [
    "data.month.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "KVJt5r2RCG8P",
    "outputId": "50b93227-8e15-44fe-bfad-2096b929673d"
   },
   "outputs": [],
   "source": [
    "data['Logarea'].hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "j3Huqz3yCG8V",
    "outputId": "056a167f-3c12-46c0-bc01-f991e9d2a259"
   },
   "outputs": [],
   "source": [
    "data.corr()['Logarea'].abs().sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "B-lOmlmhCG8c",
    "outputId": "db57556c-1fd5-4e68-ade2-160a79f09041"
   },
   "outputs": [],
   "source": [
    "data.boxplot(column='Logarea',by='DMC')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 68
    },
    "colab_type": "code",
    "id": "-pWMhgA8kvX1",
    "outputId": "ce652c36-be23-402e-b69c-33d6a9ed958d"
   },
   "outputs": [],
   "source": [
    "model = DecisionTreeRegressor()\n",
    "rfe = RFECV(model, step=0.1, cv=10)\n",
    "fit = rfe.fit(data_upsampled.drop(['area', 'Logarea_bins', 'Logarea'], axis=1), data_upsampled['Logarea'])\n",
    "\n",
    "print(\"Number of Features: \", fit.n_features_)\n",
    "print(\"Selected Features: \", fit.support_)\n",
    "print(\"Feature Ranking: \", fit.ranking_) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 105
    },
    "colab_type": "code",
    "id": "Vs1m_CFulXLV",
    "outputId": "69866001-a4dd-4cf5-8e70-80a9e619e4f4"
   },
   "outputs": [],
   "source": [
    "param_grid_dtr = {\n",
    "    'max_depth': [3, 4, 5, 6, 7, 8],\n",
    "    'max_features': [2, 3, 4, 5, 6, 7],\n",
    "    'random_state': [123],\n",
    "    'min_weight_fraction_leaf': [0, 0.3, 0.5],\n",
    "    'min_samples_split': [2, 3, 4, 5],\n",
    "    'min_samples_leaf': [3, 5, 7, 9]\n",
    "}\n",
    "\n",
    "dtr = DecisionTreeRegressor()\n",
    "dtrc = GridSearchCV(dtr, param_grid_dtr, scoring='neg_mean_squared_error', cv=10)\n",
    "dtrc.fit(X_train, y_train)\n",
    "print(dtrc.best_score_)\n",
    "print(dtrc.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "B-rdaozFrqgZ"
   },
   "outputs": [],
   "source": [
    "param_grid_abr = {\n",
    "    'n_estimators': [65, 90, 120, 200],\n",
    "    'learning_rate': [1, 0.3, 0.1, 0.01],\n",
    "    'loss': ['linear', 'square', 'exponential'],\n",
    "    'random_state': [0]\n",
    "}\n",
    "\n",
    "abr = AdaBoostRegressor()\n",
    "abrc = GridSearchCV(abr, param_grid_abr, scoring='neg_mean_squared_error', cv=10)\n",
    "abrc.fit(X_train, y_train)\n",
    "print(abrc.best_score_)\n",
    "print(abrc.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 68
    },
    "colab_type": "code",
    "id": "ZMVEswpOrqYj",
    "outputId": "c59fcf0e-e68c-4cdb-d72b-1e6f3724f37e"
   },
   "outputs": [],
   "source": [
    "abr = AdaBoostRegressor(n_estimators=200, learning_rate=1, loss='square', random_state=0)\n",
    "\n",
    "abr.fit(X_train, y_train)\n",
    "\n",
    "y_predict = abr.predict(X_test)\n",
    "y_true = y_test\n",
    "\n",
    "mse = mean_squared_error(y_true, y_predict)\n",
    "rmse = np.sqrt(mse)\n",
    "print('RMSE: ', rmse.round(3))\n",
    "print('MAD: ', mean_absolute_deviation(y_predict).round(3))\n",
    "print('NLL: ', get_nll(y_true, y_predict).round(3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 507
    },
    "colab_type": "code",
    "id": "M1praV9BCG8x",
    "outputId": "6ac01c11-5a7c-4117-92ea-11aa65dc15a8"
   },
   "outputs": [],
   "source": [
    "corr_matrix = X_train.corr(method='spearman')\n",
    "\n",
    "plt.figure(figsize=(12, 8))\n",
    "sns.heatmap(corr_matrix, cmap='YlGnBu')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 296
    },
    "colab_type": "code",
    "id": "76yJnzUFCG9V",
    "outputId": "a0912182-5ee9-496f-a0ff-5dcc63221d60"
   },
   "outputs": [],
   "source": [
    "sns.distplot(data_upsampled['Logarea'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "e83EUhqdWtB2"
   },
   "outputs": [],
   "source": [
    "sns.pairplot(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Q1i7nuqSCG9b",
    "outputId": "d1b6f6fa-cc39-47ff-fce5-2aa2971b2159"
   },
   "outputs": [],
   "source": [
    "plt.scatter(data['Y'], data['Logarea'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "zYFGueX1CG9e",
    "outputId": "c46cfc77-e0d4-4897-f72b-cbcbceac7d04"
   },
   "outputs": [],
   "source": [
    "data.boxplot(column='Logarea', by='temp')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "h3ch5wO2CG9h",
    "outputId": "7fcb6f30-2df1-4906-ee7b-a65998bacc1a"
   },
   "outputs": [],
   "source": [
    "data.hist(bins=50, figsize=(20,12), ec='w')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 105
    },
    "colab_type": "code",
    "id": "oHA4B-0NCG92",
    "outputId": "9a3ae27f-a248-4b45-ba65-ed6b09688889"
   },
   "outputs": [],
   "source": [
    "param_grid_dtr = {\n",
    "    'max_depth': [5, 6, 7, 8, 9],\n",
    "    'max_features': [3, 4, 5, 6],\n",
    "    'random_state': [0],\n",
    "    'min_weight_fraction_leaf': [0, 0.3, 0.5],\n",
    "    'min_samples_split': [2, 3, 4, 5],\n",
    "    'min_samples_leaf': [2, 3, 5]\n",
    "}\n",
    "\n",
    "dtr = DecisionTreeRegressor()\n",
    "dtrc = GridSearchCV(dtr, param_grid_dtr, scoring='neg_mean_squared_error', cv=10)\n",
    "dtrc.fit(X_train, y_train)\n",
    "print(dtrc.best_score_)\n",
    "print(dtrc.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "WMNTu4YSCG9l",
    "outputId": "cf17df80-16ad-442c-8cdb-eb5d61c28fda",
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "dot_data = StringIO()\n",
    "export_graphviz(dtr, out_file=dot_data, special_characters=True, filled=False, rounded=True, \n",
    "                rotate=True, feature_names = X_train.columns)\n",
    "graph = pydotplus.graph_from_dot_data(dot_data.getvalue())  \n",
    "graph.write_png('Decision Tree Regressor.png')\n",
    "Image(graph.create_png())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "N3WNT4RhCG9p",
    "outputId": "80e2e878-f0db-4724-a2ce-2f12c8ca20d9"
   },
   "outputs": [],
   "source": [
    "dtr.n_features_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "6YMLD8d2CG9r",
    "outputId": "992d1034-8c7c-4087-fc86-47e1e5b6e2a8"
   },
   "outputs": [],
   "source": [
    "dtr.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "DjU1FZySCG9t",
    "outputId": "01193bd8-964a-424b-d160-e9c0dee49cfd"
   },
   "outputs": [],
   "source": [
    "dtr.get_depth()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "CzlPwxcVCG9v",
    "outputId": "1ec8dd1c-f162-43b3-b7b6-acbe490b9e82"
   },
   "outputs": [],
   "source": [
    "dtr.get_n_leaves()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "FlgSQoh5CG9y",
    "outputId": "c8ab86cf-4ac2-4591-a916-c9e1eb86546f"
   },
   "outputs": [],
   "source": [
    "dtr.max_features_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "9hnKqOs9mlCw"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "dJl4OJxEmz_r"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 513
    },
    "colab_type": "code",
    "id": "pZBzR6w2tIm3",
    "outputId": "3144880b-4c10-42a4-9b9a-6fbc0b38ec44"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "JWjJgx1QtsZI"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "Forest Fire Decision Tree.ipynb",
   "provenance": []
  },
  "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}