clustering_approach/kmeans+svm+undersample.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this approach we will make almost everything the same as in kmeans+svm\n",
    "\n",
    "Except for the fact that we will undersample our data, so there will be no class imbalance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "%matplotlib inline\n",
    "np.random.seed(59)\n",
    "\n",
    "# set the max columns to none\n",
    "pd.set_option('display.max_columns', None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "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>account_length</th>\n",
       "      <th>international_plan</th>\n",
       "      <th>voice_mail_plan</th>\n",
       "      <th>number_vmail_messages</th>\n",
       "      <th>total_intl_minutes</th>\n",
       "      <th>total_intl_calls</th>\n",
       "      <th>total_intl_charge</th>\n",
       "      <th>number_customer_service_calls</th>\n",
       "      <th>churn</th>\n",
       "      <th>total_minutes</th>\n",
       "      <th>total_calls</th>\n",
       "      <th>total_charge</th>\n",
       "      <th>area_code_area_code_408</th>\n",
       "      <th>area_code_area_code_415</th>\n",
       "      <th>area_code_area_code_510</th>\n",
       "      <th>state_MD</th>\n",
       "      <th>state_MN</th>\n",
       "      <th>state_NJ</th>\n",
       "      <th>state_TX</th>\n",
       "      <th>state_WV</th>\n",
       "      <th>state_other</th>\n",
       "      <th>avg_mt_minutes</th>\n",
       "      <th>avg_mt_calls</th>\n",
       "      <th>avg_mt_charge</th>\n",
       "      <th>avg_call_charge</th>\n",
       "      <th>avg_intl_call_charge</th>\n",
       "      <th>avg_call_minutes</th>\n",
       "      <th>avg_intl_call_minutes</th>\n",
       "      <th>both_plans</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>107</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>26</td>\n",
       "      <td>13.7</td>\n",
       "      <td>3</td>\n",
       "      <td>3.70</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>611.5</td>\n",
       "      <td>329</td>\n",
       "      <td>55.54</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5.714953</td>\n",
       "      <td>3.074766</td>\n",
       "      <td>0.519065</td>\n",
       "      <td>0.168815</td>\n",
       "      <td>1.233333</td>\n",
       "      <td>1.858663</td>\n",
       "      <td>4.566667</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>137</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>12.2</td>\n",
       "      <td>5</td>\n",
       "      <td>3.29</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>527.2</td>\n",
       "      <td>328</td>\n",
       "      <td>59.00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.848175</td>\n",
       "      <td>2.394161</td>\n",
       "      <td>0.430657</td>\n",
       "      <td>0.179878</td>\n",
       "      <td>0.658000</td>\n",
       "      <td>1.607317</td>\n",
       "      <td>2.440000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>84</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>6.6</td>\n",
       "      <td>7</td>\n",
       "      <td>1.78</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>558.2</td>\n",
       "      <td>248</td>\n",
       "      <td>65.02</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>6.645238</td>\n",
       "      <td>2.952381</td>\n",
       "      <td>0.774048</td>\n",
       "      <td>0.262177</td>\n",
       "      <td>0.254286</td>\n",
       "      <td>2.250806</td>\n",
       "      <td>0.942857</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>75</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.1</td>\n",
       "      <td>3</td>\n",
       "      <td>2.73</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>501.9</td>\n",
       "      <td>356</td>\n",
       "      <td>49.36</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>6.692000</td>\n",
       "      <td>4.746667</td>\n",
       "      <td>0.658133</td>\n",
       "      <td>0.138652</td>\n",
       "      <td>0.910000</td>\n",
       "      <td>1.409831</td>\n",
       "      <td>3.366667</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>121</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>24</td>\n",
       "      <td>7.5</td>\n",
       "      <td>7</td>\n",
       "      <td>2.03</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>779.3</td>\n",
       "      <td>314</td>\n",
       "      <td>76.28</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>6.440496</td>\n",
       "      <td>2.595041</td>\n",
       "      <td>0.630413</td>\n",
       "      <td>0.242930</td>\n",
       "      <td>0.290000</td>\n",
       "      <td>2.481847</td>\n",
       "      <td>1.071429</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   account_length  international_plan  voice_mail_plan  number_vmail_messages  \\\n",
       "0             107                   0                1                     26   \n",
       "1             137                   0                0                      0   \n",
       "2              84                   1                0                      0   \n",
       "3              75                   1                0                      0   \n",
       "4             121                   0                1                     24   \n",
       "\n",
       "   total_intl_minutes  total_intl_calls  total_intl_charge  \\\n",
       "0                13.7                 3               3.70   \n",
       "1                12.2                 5               3.29   \n",
       "2                 6.6                 7               1.78   \n",
       "3                10.1                 3               2.73   \n",
       "4                 7.5                 7               2.03   \n",
       "\n",
       "   number_customer_service_calls  churn  total_minutes  total_calls  \\\n",
       "0                              1      0          611.5          329   \n",
       "1                              0      0          527.2          328   \n",
       "2                              2      0          558.2          248   \n",
       "3                              3      0          501.9          356   \n",
       "4                              3      0          779.3          314   \n",
       "\n",
       "   total_charge  area_code_area_code_408  area_code_area_code_415  \\\n",
       "0         55.54                        0                        1   \n",
       "1         59.00                        0                        1   \n",
       "2         65.02                        1                        0   \n",
       "3         49.36                        0                        1   \n",
       "4         76.28                        0                        0   \n",
       "\n",
       "   area_code_area_code_510  state_MD  state_MN  state_NJ  state_TX  state_WV  \\\n",
       "0                        0         0         0         0         0         0   \n",
       "1                        0         0         0         1         0         0   \n",
       "2                        0         0         0         0         0         0   \n",
       "3                        0         0         0         0         0         0   \n",
       "4                        1         0         0         0         0         0   \n",
       "\n",
       "   state_other  avg_mt_minutes  avg_mt_calls  avg_mt_charge  avg_call_charge  \\\n",
       "0            1        5.714953      3.074766       0.519065         0.168815   \n",
       "1            0        3.848175      2.394161       0.430657         0.179878   \n",
       "2            1        6.645238      2.952381       0.774048         0.262177   \n",
       "3            1        6.692000      4.746667       0.658133         0.138652   \n",
       "4            1        6.440496      2.595041       0.630413         0.242930   \n",
       "\n",
       "   avg_intl_call_charge  avg_call_minutes  avg_intl_call_minutes  both_plans  \n",
       "0              1.233333          1.858663               4.566667           0  \n",
       "1              0.658000          1.607317               2.440000           0  \n",
       "2              0.254286          2.250806               0.942857           0  \n",
       "3              0.910000          1.409831               3.366667           0  \n",
       "4              0.290000          2.481847               1.071429           0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from preprocessing import *\n",
    "\n",
    "path = 'data\\\\train.csv'\n",
    "\n",
    "df = pd.read_csv(path)\n",
    "df = convert_data(df)\n",
    "df = engineer_features(df)\n",
    "df = select_features(df)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Shuffle the Dataset\n",
    "shuffled_df = df.sample(frac=1, random_state=59)\n",
    "\n",
    "# Put all the fraud class in a separate dataset\n",
    "churn_df = shuffled_df.loc[shuffled_df['churn'] == 1]\n",
    "\n",
    "#Randomly select 492 observations from the non-fraud (majority class)\n",
    "non_churn_df = shuffled_df.loc[shuffled_df['churn'] == 0].sample(n=churn_df.shape[0], random_state=59)\n",
    "\n",
    "# Concatenate both dataframes again\n",
    "normalized_df = pd.concat([churn_df, non_churn_df])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the dataset after the undersampling\n",
    "plt.figure(figsize=(8, 4))\n",
    "sns.set_style('white')\n",
    "sns.countplot(x='churn', data=normalized_df)\n",
    "plt.title('Balanced Classes')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = normalized_df.drop('churn', axis=1)\n",
    "y = normalized_df['churn']\n",
    "\n",
    "X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2,\n",
    "                                                  random_state=59)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "st_scaler = StandardScaler()\n",
    "\n",
    "X_train_st = st_scaler.fit_transform(X_train)\n",
    "X_val_st = st_scaler.transform(X_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "km_list = list()\n",
    "\n",
    "for i in range(1, 30):\n",
    "    km = KMeans(n_clusters=i, init='k-means++', max_iter=300, n_init=10, random_state=59)\n",
    "    km = km.fit(X_train_st)\n",
    "    \n",
    "    km_list.append(pd.Series({'clusters': i, \n",
    "                              'inertia': km.inertia_,\n",
    "                              'model': km}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "k = pd.concat(km_list, axis=1).T[['clusters','inertia']]\n",
    "\n",
    "# Visualize\n",
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "fig.patch.set_facecolor('white')\n",
    "\n",
    "plt.plot(k['clusters'], k['inertia'], 'bo-')\n",
    "\n",
    "# Remove ticks\n",
    "ax.xaxis.set_ticks_position('none')\n",
    "ax.yaxis.set_ticks_position('none')\n",
    "\n",
    "# Remove axes splines\n",
    "for i in ['top','right']:\n",
    "    ax.spines[i].set_visible(False)\n",
    "\n",
    "ax.set_xticks(range(0, 30, 2))\n",
    "ax.set(xlabel='Cluster', ylabel='Inertia')\n",
    "\n",
    "plt.suptitle('The Elbow Method: Optimal Number of Clusters', size=26)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "13 is enough"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "kmeans = KMeans(n_clusters=13, max_iter=300, n_init=10, random_state=59)\n",
    "kmeans.fit(X_train_st)\n",
    "\n",
    "predicted_clusters_train = kmeans.predict(X_train_st)\n",
    "X_train_st = np.column_stack((X_train_st, predicted_clusters_train))\n",
    "\n",
    "predicted_clusters_val = kmeans.predict(X_val_st)\n",
    "X_val_st = np.column_stack((X_val_st, predicted_clusters_val))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "svm = SVC()\n",
    "svm.fit(X_train_st, y_train)\n",
    "y_pred = svm.predict(X_val_st)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.80      0.87      0.83       121\n",
      "           1       0.85      0.78      0.82       119\n",
      "\n",
      "    accuracy                           0.82       240\n",
      "   macro avg       0.83      0.82      0.82       240\n",
      "weighted avg       0.83      0.82      0.82       240\n",
      "\n",
      "F1 Score 0.8157894736842106\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report, f1_score\n",
    "\n",
    "print(classification_report(y_val, y_pred))\n",
    "print('F1 Score', f1_score(y_val, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "F1 Score = 0.816"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we will try something more sophisticated like SMOTE (for oversampling instead of undersampling)"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "e92dda350e586b0c334f0f5e12c30469449e88a6ff79a1f1c33a5faa01dbdb98"
  },
  "kernelspec": {
   "display_name": "Python 3.10.0 64-bit",
   "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": "3.10.0"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}