04-SMOTE.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 5G UE-VBS Selection ML Model\n",
    "- <strong>Environment Setup:</strong> import libraries and load the dataset\n",
    "- <strong>Imbalanced Data Analysis:</strong> understanding how the dataset balance is restored using random under sampling\n",
    "- <strong>SMOTE:</strong> Creating synthetic data for the minority class using SMOTE(Synthetic Minority Over-sampling Technique), takes random samples from the minority class, finds its nearest k neighbors, and then selects a point between the randomly selected data point and its nearest k neighbors to generate synthetic data\n",
    "- <strong>Inference:</strong> understand how the model produces results and choosing the best classifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Environmnet Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "jl27hA_Ucd9N"
   },
   "outputs": [],
   "source": [
    "# Pre-requisites\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "# Store the classifier models to save time\n",
    "import joblib\n",
    "\n",
    "# Preprocessing\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# Classifiers from scikit-learn\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "\n",
    "# Performance metrics\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import confusion_matrix\n",
    "# from sklearn.metrics import ConfusionMatrixDisplay\n",
    "from sklearn.metrics import plot_confusion_matrix\n",
    "\n",
    "from sklearn.metrics import roc_curve\n",
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "# Model Selection\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "# Imbalanced dataset helpers\n",
    "from imblearn.over_sampling import SMOTE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "0HA4SPpocpEP"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>distance_from_base_station</th>\n",
       "      <th>battery_discharge_rate</th>\n",
       "      <th>power</th>\n",
       "      <th>memory</th>\n",
       "      <th>cpu_power</th>\n",
       "      <th>qos</th>\n",
       "      <th>qoe</th>\n",
       "      <th>data_rate</th>\n",
       "      <th>eligibility</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>465.575800</td>\n",
       "      <td>45.312870</td>\n",
       "      <td>9.301212</td>\n",
       "      <td>6</td>\n",
       "      <td>65.953269</td>\n",
       "      <td>0.992467</td>\n",
       "      <td>0.147388</td>\n",
       "      <td>3.237754e+04</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>383.632002</td>\n",
       "      <td>107.164485</td>\n",
       "      <td>10.982724</td>\n",
       "      <td>12</td>\n",
       "      <td>48.639950</td>\n",
       "      <td>0.993799</td>\n",
       "      <td>0.136998</td>\n",
       "      <td>4.545044e+05</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>178.448694</td>\n",
       "      <td>42.099486</td>\n",
       "      <td>17.630753</td>\n",
       "      <td>6</td>\n",
       "      <td>88.386756</td>\n",
       "      <td>0.995114</td>\n",
       "      <td>0.125348</td>\n",
       "      <td>4.774090e+04</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>519.320506</td>\n",
       "      <td>117.518415</td>\n",
       "      <td>8.352311</td>\n",
       "      <td>8</td>\n",
       "      <td>93.333433</td>\n",
       "      <td>0.992822</td>\n",
       "      <td>0.143667</td>\n",
       "      <td>1.416554e+06</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>643.812013</td>\n",
       "      <td>67.063529</td>\n",
       "      <td>6.485839</td>\n",
       "      <td>4</td>\n",
       "      <td>30.084433</td>\n",
       "      <td>0.992436</td>\n",
       "      <td>0.148444</td>\n",
       "      <td>5.775633e+04</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   distance_from_base_station  battery_discharge_rate      power  memory  \\\n",
       "0                  465.575800               45.312870   9.301212       6   \n",
       "1                  383.632002              107.164485  10.982724      12   \n",
       "2                  178.448694               42.099486  17.630753       6   \n",
       "3                  519.320506              117.518415   8.352311       8   \n",
       "4                  643.812013               67.063529   6.485839       4   \n",
       "\n",
       "   cpu_power       qos       qoe     data_rate  eligibility  \n",
       "0  65.953269  0.992467  0.147388  3.237754e+04            0  \n",
       "1  48.639950  0.993799  0.136998  4.545044e+05            0  \n",
       "2  88.386756  0.995114  0.125348  4.774090e+04            0  \n",
       "3  93.333433  0.992822  0.143667  1.416554e+06            0  \n",
       "4  30.084433  0.992436  0.148444  5.775633e+04            0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Loading the dataset\n",
    "dataset_location = './Data/encoded_master_simulation_1000_2500.csv'\n",
    "df = pd.read_csv(dataset_location, index_col=0)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 5250 entries, 0 to 5249\n",
      "Data columns (total 9 columns):\n",
      " #   Column                      Non-Null Count  Dtype  \n",
      "---  ------                      --------------  -----  \n",
      " 0   distance_from_base_station  5250 non-null   float64\n",
      " 1   battery_discharge_rate      5250 non-null   float64\n",
      " 2   power                       5250 non-null   float64\n",
      " 3   memory                      5250 non-null   int64  \n",
      " 4   cpu_power                   5250 non-null   float64\n",
      " 5   qos                         5250 non-null   float64\n",
      " 6   qoe                         5250 non-null   float64\n",
      " 7   data_rate                   5250 non-null   float64\n",
      " 8   eligibility                 5250 non-null   int64  \n",
      "dtypes: float64(7), int64(2)\n",
      "memory usage: 410.2 KB\n",
      "None\n",
      "\n",
      "Shape: (5250, 9)\n"
     ]
    }
   ],
   "source": [
    "# Understanding the column data\n",
    "print(df.info())\n",
    "print(f\"\\nShape: {df.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dataset Without Handling Imbalance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0    4243\n",
      "1    1007\n",
      "Name: eligibility, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# Plotting the percentage of observations that fall under each class\n",
    "ax = df[\"eligibility\"].value_counts().sort_values().plot(kind=\"barh\", color=[\"r\", \"g\"])\n",
    "totals= []\n",
    "for i in ax.patches:\n",
    "    totals.append(i.get_width())\n",
    "total = sum(totals)\n",
    "for i in ax.patches:\n",
    "     ax.text(i.get_width()+.3, i.get_y()+.20, \n",
    "     str(round((i.get_width()/total)*100, 2))+'%', \n",
    "     fontsize=10, color='black')\n",
    "plt.title(\"Eligibility\", fontsize=20)\n",
    "plt.xlabel(\"Count\", fontsize=14)\n",
    "plt.ylabel(\"Class\", fontsize=14)\n",
    "plt.show()\n",
    "print(df[\"eligibility\"].value_counts())\n",
    "fig = ax.get_figure()\n",
    "fig.savefig(\"Fig/04-01-Class_Imbalance.png\")\n",
    "fig.savefig(\"Eps/04-01-Class_Imbalance.eps\", format='eps')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Dataset shape: (5250, 9)\n",
      "features_train shape: (3675, 8) | features_test shape: (1575, 8)\n",
      "labels_train shape: (3675,) | labels_test shape: (1575,)\n"
     ]
    }
   ],
   "source": [
    "features = df.loc[:, df.columns != \"eligibility\"]\n",
    "labels = np.asarray(df[\"eligibility\"])\n",
    "# Number of samples/observations/rows is greater than 100,000\n",
    "print(f\"\\nDataset shape: {df.shape}\")\n",
    "features_train, features_test, labels_train, labels_test = train_test_split(features, labels, test_size=0.30, random_state=42)\n",
    "print(f\"features_train shape: {features_train.shape} | features_test shape: {features_test.shape}\")\n",
    "print(f\"labels_train shape: {labels_train.shape} | labels_test shape: {labels_test.shape}\")\n",
    "# print(f\"{features_train.shape[1]} Features: {features_train.columns.to_list()}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training Data Without Handling Imbalance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0    2962\n",
      "1     713\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# Plotting the percentage of observations that fall under each class\n",
    "ax = pd.DataFrame(labels_train).value_counts().sort_values().plot(kind=\"barh\", color=[\"r\", \"g\"])\n",
    "ax.set_axisbelow(True)\n",
    "ax.grid()\n",
    "totals= []\n",
    "for i in ax.patches:\n",
    "    totals.append(i.get_width())\n",
    "total = sum(totals)\n",
    "for i in ax.patches:\n",
    "     ax.text(i.get_width()+.3, i.get_y()+.20, \n",
    "     str(round((i.get_width()/total)*100, 2))+'%', \n",
    "     fontsize=10, color='black')\n",
    "plt.title(\"Eligibility\", fontsize=20)\n",
    "plt.xlabel(\"Count\", fontsize=14)\n",
    "plt.ylabel(\"Class\", fontsize=14)\n",
    "plt.show()\n",
    "print(pd.DataFrame(labels_train).value_counts())\n",
    "fig = ax.get_figure()\n",
    "fig.savefig(\"Fig/04-02-Training_Class_Imbalance.png\")\n",
    "fig.savefig(\"Eps/04-02-Training_Class_Imbalance.eps\", format='eps')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SMOTE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "smote = SMOTE(random_state=42)\n",
    "features_train, labels_train = smote.fit_resample(features_train, labels_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# SMOTE takes too long for large datasets\n",
    "features_train.to_csv(\"./Data/smote_features_train.csv\")\n",
    "pd.DataFrame(labels_train).to_csv(\"./Data/smote_labels_train.csv\")\n",
    "features_train = pd.read_csv(\"./Data/smote_features_train.csv\", index_col=0)\n",
    "labels_train = pd.read_csv(\"./Data/smote_labels_train.csv\", index_col=0)\n",
    "# labels_train.shape\n",
    "labels_train = labels_train[\"0\"].values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Balancing the Training Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1    2962\n",
      "0    2962\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# Plotting the percentage of observations that fall under each class\n",
    "ax = pd.DataFrame(labels_train).value_counts().sort_values().plot(kind=\"barh\", color=[\"r\", \"g\"])\n",
    "ax.set_axisbelow(True)\n",
    "ax.grid()\n",
    "totals= []\n",
    "for i in ax.patches:\n",
    "    totals.append(i.get_width())\n",
    "total = sum(totals)\n",
    "for i in ax.patches:\n",
    "     ax.text(i.get_width()+.3, i.get_y()+.20, \n",
    "     str(round((i.get_width()/total)*100, 2))+'%', \n",
    "     fontsize=10, color='black')\n",
    "plt.title(\"Eligibility\", fontsize=20)\n",
    "plt.xlabel(\"Count\", fontsize=14)\n",
    "plt.ylabel(\"Class\", fontsize=14)\n",
    "plt.show()\n",
    "print(pd.DataFrame(labels_train).value_counts())\n",
    "fig = ax.get_figure()\n",
    "fig.savefig(\"Fig/04-03-Balanced.png\")\n",
    "fig.savefig(\"Eps/04-03-Balanced.eps\", format='eps')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Modeling Different Classifiers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      0.96      0.97      1281\n",
      "           1       0.85      0.91      0.88       294\n",
      "\n",
      "    accuracy                           0.95      1575\n",
      "   macro avg       0.91      0.94      0.92      1575\n",
      "weighted avg       0.95      0.95      0.95      1575\n",
      "\n",
      "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay object at 0x000001642B88CEB8>\n",
      "\n",
      "Cross-Validation cv=5\n",
      "0.95 0.96 0.98 0.96 0.98 \n",
      "\n",
      "Mean: 0.96\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = RandomForestClassifier(n_jobs=-1, random_state=42)\n",
    "model.fit(features_train, labels_train)\n",
    "joblib.dump(model, \"./SMOTE/RandomForestClassifier.joblib\")\n",
    "model = joblib.load(\"./SMOTE/RandomForestClassifier.joblib\")\n",
    "model_pred = model.predict(features_test)\n",
    "# print(confusion_matrix(labels_test, model_pred))\n",
    "print(classification_report(labels_test, model_pred))\n",
    "# conf_mat_plot = ConfusionMatrixDisplay.from_estimator(model, features_test, labels_test)\n",
    "conf_mat_plot = plot_confusion_matrix(model, features_test, labels_test)\n",
    "plt.title(\"Random Forest Classifier\")\n",
    "plt.savefig(\"Fig/04-04-RFC_Confusion_Matrix_SMOTE.png\")\n",
    "plt.savefig(\"Eps/04-04-RFC_Confusion_Matrix_SMOTE.eps\", format='eps')\n",
    "print(conf_mat_plot)\n",
    "print(\"\\nCross-Validation cv=5\")\n",
    "scores = cross_val_score(model, features_train, labels_train, cv=5)\n",
    "for score in scores:\n",
    "    print(\"%.2f\" %score, end=\" \")\n",
    "print(\"\\n\\nMean: %.2f\" %scores.mean())\n",
    "del model\n",
    "del model_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gradient Boosting Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      0.96      0.97      1281\n",
      "           1       0.83      0.94      0.88       294\n",
      "\n",
      "    accuracy                           0.95      1575\n",
      "   macro avg       0.91      0.95      0.92      1575\n",
      "weighted avg       0.96      0.95      0.95      1575\n",
      "\n",
      "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay object at 0x000001642CD31B38>\n",
      "\n",
      "Cross-Validation cv=5\n",
      "0.94 0.95 0.96 0.96 0.97 \n",
      "\n",
      "Mean: 0.96\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = GradientBoostingClassifier(random_state=42)\n",
    "model.fit(features_train, labels_train)\n",
    "joblib.dump(model, \"./SMOTE/GradientBoostingClassifier.joblib\")\n",
    "model = joblib.load(\"./SMOTE/GradientBoostingClassifier.joblib\")\n",
    "model_pred = model.predict(features_test)\n",
    "# print(confusion_matrix(labels_test, model_pred))\n",
    "print(classification_report(labels_test, model_pred))\n",
    "# conf_mat_plot = ConfusionMatrixDisplay.from_estimator(model, features_test, labels_test)\n",
    "conf_mat_plot = plot_confusion_matrix(model, features_test, labels_test)\n",
    "plt.title(\"Gradient Boosting Classifier\")\n",
    "plt.savefig(\"Fig/04-05-GBC_Confusion_Matrix_SMOTE.png\")\n",
    "plt.savefig(\"Eps/04-05-GBC_Confusion_Matrix_SMOTE.eps\", format='eps')\n",
    "print(conf_mat_plot)\n",
    "print(\"\\nCross-Validation cv=5\")\n",
    "scores = cross_val_score(model, features_train, labels_train, cv=5)\n",
    "for score in scores:\n",
    "    print(\"%.2f\" %score, end=\" \")\n",
    "print(\"\\n\\nMean: %.2f\" %scores.mean())\n",
    "del model\n",
    "del model_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### AdaBoost Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      0.95      0.97      1281\n",
      "           1       0.82      0.90      0.86       294\n",
      "\n",
      "    accuracy                           0.94      1575\n",
      "   macro avg       0.90      0.93      0.91      1575\n",
      "weighted avg       0.95      0.94      0.95      1575\n",
      "\n",
      "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay object at 0x000001642B665CC0>\n",
      "\n",
      "Cross-Validation cv=5\n",
      "0.91 0.95 0.95 0.95 0.96 \n",
      "\n",
      "Mean: 0.94\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = AdaBoostClassifier(random_state=42)\n",
    "model.fit(features_train, labels_train)\n",
    "joblib.dump(model, \"./SMOTE/AdaBoostClassifier.joblib\")\n",
    "model = joblib.load(\"./SMOTE/AdaBoostClassifier.joblib\")\n",
    "model_pred = model.predict(features_test)\n",
    "# print(confusion_matrix(labels_test, model_pred))\n",
    "print(classification_report(labels_test, model_pred))\n",
    "# conf_mat_plot = ConfusionMatrixDisplay.from_estimator(model, features_test, labels_test)\n",
    "conf_mat_plot = plot_confusion_matrix(model, features_test, labels_test)\n",
    "plt.title(\"AdaBoost Classifier\")\n",
    "plt.savefig(\"Fig/04-06-AdaBoost_Confusion_Matrix_SMOTE.png\")\n",
    "plt.savefig(\"Eps/04-06-AdaBoost_Confusion_Matrix_SMOTE.eps\", format='eps')\n",
    "print(conf_mat_plot)\n",
    "print(\"\\nCross-Validation cv=5\")\n",
    "scores = cross_val_score(model, features_train, labels_train, cv=5)\n",
    "for score in scores:\n",
    "    print(\"%.2f\" %score, end=\" \")\n",
    "print(\"\\n\\nMean: %.2f\" %scores.mean())\n",
    "del model\n",
    "del model_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Performance Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Setup\n",
    "perf_df = pd.DataFrame(columns=[\"classifiers\", \"fpr\", \"tpr\", \"auc\"])\n",
    "\n",
    "classifiers = [\"AdaBoostClassifier\", \"GradientBoostingClassifier\", \"RandomForestClassifier\"]\n",
    "for clf in classifiers:\n",
    "    # Load the trained classifiers\n",
    "    model = joblib.load(f\"./SMOTE/{clf}.joblib\")\n",
    "    model_pred = model.predict(features_test)\n",
    "    fpr, tpr, _ = roc_curve(labels_test,  model_pred)\n",
    "    auc = roc_auc_score(labels_test, model_pred)\n",
    "    # Record the results\n",
    "    perf_df = perf_df.append({\"classifiers\": clf,\n",
    "                                        \"fpr\": fpr, \n",
    "                                        \"tpr\": tpr, \n",
    "                                        \"auc\": auc}, ignore_index=True)\n",
    "    del model\n",
    "    del model_pred\n",
    "\n",
    "# Set name of the classifiers as index labels\n",
    "perf_df.set_index(\"classifiers\", inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n",
      "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8,6), dpi=80)\n",
    "for clf_name in perf_df.index:\n",
    "    plt.plot(perf_df.loc[clf_name][\"fpr\"], \n",
    "             perf_df.loc[clf_name][\"tpr\"], \n",
    "             label=\"{}, AUC={:.2f}\".format(clf_name, perf_df.loc[clf_name][\"auc\"]))\n",
    "    \n",
    "plt.plot([0,1], [0,1], color='red', linestyle='--')\n",
    "\n",
    "plt.xticks(np.arange(0.0, 1.1, step=0.1))\n",
    "plt.xlabel(\"False Positive Rate\", fontsize=15)\n",
    "\n",
    "plt.yticks(np.arange(0.0, 1.1, step=0.1))\n",
    "plt.ylabel(\"True Positive Rate\", fontsize=15)\n",
    "\n",
    "plt.title(\"ROC AUC Analysis\", fontweight=\"bold\", fontsize=15)\n",
    "plt.legend(prop={\"size\":13}, loc=\"lower right\")\n",
    "\n",
    "plt.savefig(\"Fig/04-07-ROCAUC_After_SMOTE.png\")\n",
    "plt.savefig(\"Eps/04-07-ROCAUC_After_SMOTE.eps\", format='eps')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n",
      "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "markers = 'oD*'\n",
    "m = 0\n",
    "fig = plt.figure(figsize=(8,6), dpi=80)\n",
    "for clf_name in perf_df.index:\n",
    "    plt.plot(perf_df.loc[clf_name][\"fpr\"], \n",
    "             perf_df.loc[clf_name][\"tpr\"], marker=markers[m], markersize='5',\n",
    "             label=\"{}, AUC={:.2f}\".format(clf_name, perf_df.loc[clf_name][\"auc\"]))\n",
    "    m += 1\n",
    "    \n",
    "plt.plot([0,1], [0,1], color='red', linestyle='--')\n",
    "\n",
    "plt.xticks(np.arange(0.0, 1.1, step=0.1))\n",
    "plt.xlabel(\"False Ineligible Rate (FIR)\", fontsize=15)\n",
    "\n",
    "plt.yticks(np.arange(0.0, 1.1, step=0.1))\n",
    "plt.ylabel(\"True Eligible Rate (TER)\", fontsize=15)\n",
    "\n",
    "plt.title(\"ROC AUC Analysis\", fontweight=\"bold\", fontsize=15)\n",
    "plt.legend(prop={\"size\":13}, loc=\"lower right\")\n",
    "\n",
    "plt.savefig(\"Fig/04-08-ROCAUC_After_SMOTE.png\")\n",
    "plt.savefig(\"Eps/04-08-ROCAUC_After_SMOTE.eps\", format='eps')\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "02-Classification.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": "3.6.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}