12_handling_imbalanced_data.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "612c6c4b-d116-4358-81ee-48d3cec3f6a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from matplotlib import pyplot as plt\n",
    "import numpy as np\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "62daef70-153f-4cab-994c-4234b7b7c264",
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3fa15937-90a0-482b-9a56-9eb011129432",
   "metadata": {},
   "source": [
    "## Load the Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "126e0234-5754-4d13-b0f5-269061172e7e",
   "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>customerID</th>\n",
       "      <th>gender</th>\n",
       "      <th>SeniorCitizen</th>\n",
       "      <th>Partner</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>tenure</th>\n",
       "      <th>PhoneService</th>\n",
       "      <th>MultipleLines</th>\n",
       "      <th>InternetService</th>\n",
       "      <th>OnlineSecurity</th>\n",
       "      <th>...</th>\n",
       "      <th>DeviceProtection</th>\n",
       "      <th>TechSupport</th>\n",
       "      <th>StreamingTV</th>\n",
       "      <th>StreamingMovies</th>\n",
       "      <th>Contract</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>PaymentMethod</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6543</th>\n",
       "      <td>8050-XGRVL</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>62</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>One year</td>\n",
       "      <td>No</td>\n",
       "      <td>Credit card (automatic)</td>\n",
       "      <td>54.75</td>\n",
       "      <td>3425.35</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4652</th>\n",
       "      <td>3178-CIFOT</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>23</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>...</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>One year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Credit card (automatic)</td>\n",
       "      <td>19.65</td>\n",
       "      <td>478.1</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>821</th>\n",
       "      <td>5318-YKDPV</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>43</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Two year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Credit card (automatic)</td>\n",
       "      <td>79.15</td>\n",
       "      <td>3566.6</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6145</th>\n",
       "      <td>5010-IPEAQ</td>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>67</td>\n",
       "      <td>No</td>\n",
       "      <td>No phone service</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Two year</td>\n",
       "      <td>No</td>\n",
       "      <td>Credit card (automatic)</td>\n",
       "      <td>54.20</td>\n",
       "      <td>3623.95</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2624</th>\n",
       "      <td>0820-FNRNX</td>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>17</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>DSL</td>\n",
       "      <td>No</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>66.70</td>\n",
       "      <td>1077.05</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      customerID  gender  SeniorCitizen Partner Dependents  tenure  \\\n",
       "6543  8050-XGRVL  Female              0     Yes        Yes      62   \n",
       "4652  3178-CIFOT  Female              0      No         No      23   \n",
       "821   5318-YKDPV    Male              0     Yes        Yes      43   \n",
       "6145  5010-IPEAQ  Female              0     Yes        Yes      67   \n",
       "2624  0820-FNRNX    Male              0     Yes        Yes      17   \n",
       "\n",
       "     PhoneService     MultipleLines InternetService       OnlineSecurity  ...  \\\n",
       "6543          Yes                No             DSL                  Yes  ...   \n",
       "4652          Yes                No              No  No internet service  ...   \n",
       "821           Yes                No             DSL                  Yes  ...   \n",
       "6145           No  No phone service             DSL                  Yes  ...   \n",
       "2624          Yes               Yes             DSL                   No  ...   \n",
       "\n",
       "         DeviceProtection          TechSupport          StreamingTV  \\\n",
       "6543                  Yes                   No                   No   \n",
       "4652  No internet service  No internet service  No internet service   \n",
       "821                   Yes                  Yes                  Yes   \n",
       "6145                  Yes                  Yes                  Yes   \n",
       "2624                  Yes                  Yes                   No   \n",
       "\n",
       "          StreamingMovies        Contract PaperlessBilling  \\\n",
       "6543                   No        One year               No   \n",
       "4652  No internet service        One year              Yes   \n",
       "821                   Yes        Two year              Yes   \n",
       "6145                   No        Two year               No   \n",
       "2624                   No  Month-to-month              Yes   \n",
       "\n",
       "                PaymentMethod MonthlyCharges  TotalCharges Churn  \n",
       "6543  Credit card (automatic)          54.75       3425.35    No  \n",
       "4652  Credit card (automatic)          19.65         478.1    No  \n",
       "821   Credit card (automatic)          79.15        3566.6    No  \n",
       "6145  Credit card (automatic)          54.20       3623.95    No  \n",
       "2624             Mailed check          66.70       1077.05    No  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(\"customer_churn.csv\")\n",
    "df.sample(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "48baf362-f17f-4eb4-9a15-6998e3f4562c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "No     5174\n",
       "Yes    1869\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.Churn.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d4ad3262-2bb7-4c99-a5d9-3f645c96e1ad",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "73.46301292063042"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "517400/df.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "b92e92b2-e79f-4278-9035-00bbe1f04e49",
   "metadata": {},
   "outputs": [],
   "source": [
    "# drop unnecessary column\n",
    "df.drop(\"customerID\", axis=\"columns\", inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c7538cad-5305-46d1-91df-af77f49d0df0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "gender               object\n",
       "SeniorCitizen         int64\n",
       "Partner              object\n",
       "Dependents           object\n",
       "tenure                int64\n",
       "PhoneService         object\n",
       "MultipleLines        object\n",
       "InternetService      object\n",
       "OnlineSecurity       object\n",
       "OnlineBackup         object\n",
       "DeviceProtection     object\n",
       "TechSupport          object\n",
       "StreamingTV          object\n",
       "StreamingMovies      object\n",
       "Contract             object\n",
       "PaperlessBilling     object\n",
       "PaymentMethod        object\n",
       "MonthlyCharges      float64\n",
       "TotalCharges         object\n",
       "Churn                object\n",
       "dtype: object"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "37fb14fe-6602-417e-af8e-05c969fcc6c3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['29.85', '1889.5', '108.15', ..., '346.45', '306.6', '6844.5'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.TotalCharges.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d5d7518f-5b15-43e6-9411-546fd447f6e5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       False\n",
       "1       False\n",
       "2       False\n",
       "3       False\n",
       "4       False\n",
       "        ...  \n",
       "7038    False\n",
       "7039    False\n",
       "7040    False\n",
       "7041    False\n",
       "7042    False\n",
       "Name: TotalCharges, Length: 7043, dtype: bool"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.to_numeric(df.TotalCharges,errors='coerce').isnull()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "696354cc-3b43-4d6f-9c7c-2c5bd47a7346",
   "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>gender</th>\n",
       "      <th>SeniorCitizen</th>\n",
       "      <th>Partner</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>tenure</th>\n",
       "      <th>PhoneService</th>\n",
       "      <th>MultipleLines</th>\n",
       "      <th>InternetService</th>\n",
       "      <th>OnlineSecurity</th>\n",
       "      <th>OnlineBackup</th>\n",
       "      <th>DeviceProtection</th>\n",
       "      <th>TechSupport</th>\n",
       "      <th>StreamingTV</th>\n",
       "      <th>StreamingMovies</th>\n",
       "      <th>Contract</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>PaymentMethod</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>488</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No phone service</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Two year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Bank transfer (automatic)</td>\n",
       "      <td>52.55</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>753</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>Two year</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>20.25</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>936</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Two year</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>80.85</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1082</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>Two year</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>25.75</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1340</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No phone service</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Two year</td>\n",
       "      <td>No</td>\n",
       "      <td>Credit card (automatic)</td>\n",
       "      <td>56.05</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3331</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>Two year</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>19.85</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3826</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>Two year</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>25.35</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4380</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>Two year</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>20.00</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5218</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>One year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>19.70</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6670</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>DSL</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Two year</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>73.35</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6754</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Two year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Bank transfer (automatic)</td>\n",
       "      <td>61.90</td>\n",
       "      <td></td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      gender  SeniorCitizen Partner Dependents  tenure PhoneService  \\\n",
       "488   Female              0     Yes        Yes       0           No   \n",
       "753     Male              0      No        Yes       0          Yes   \n",
       "936   Female              0     Yes        Yes       0          Yes   \n",
       "1082    Male              0     Yes        Yes       0          Yes   \n",
       "1340  Female              0     Yes        Yes       0           No   \n",
       "3331    Male              0     Yes        Yes       0          Yes   \n",
       "3826    Male              0     Yes        Yes       0          Yes   \n",
       "4380  Female              0     Yes        Yes       0          Yes   \n",
       "5218    Male              0     Yes        Yes       0          Yes   \n",
       "6670  Female              0     Yes        Yes       0          Yes   \n",
       "6754    Male              0      No        Yes       0          Yes   \n",
       "\n",
       "         MultipleLines InternetService       OnlineSecurity  \\\n",
       "488   No phone service             DSL                  Yes   \n",
       "753                 No              No  No internet service   \n",
       "936                 No             DSL                  Yes   \n",
       "1082               Yes              No  No internet service   \n",
       "1340  No phone service             DSL                  Yes   \n",
       "3331                No              No  No internet service   \n",
       "3826               Yes              No  No internet service   \n",
       "4380                No              No  No internet service   \n",
       "5218                No              No  No internet service   \n",
       "6670               Yes             DSL                   No   \n",
       "6754               Yes             DSL                  Yes   \n",
       "\n",
       "             OnlineBackup     DeviceProtection          TechSupport  \\\n",
       "488                    No                  Yes                  Yes   \n",
       "753   No internet service  No internet service  No internet service   \n",
       "936                   Yes                  Yes                   No   \n",
       "1082  No internet service  No internet service  No internet service   \n",
       "1340                  Yes                  Yes                  Yes   \n",
       "3331  No internet service  No internet service  No internet service   \n",
       "3826  No internet service  No internet service  No internet service   \n",
       "4380  No internet service  No internet service  No internet service   \n",
       "5218  No internet service  No internet service  No internet service   \n",
       "6670                  Yes                  Yes                  Yes   \n",
       "6754                  Yes                   No                  Yes   \n",
       "\n",
       "              StreamingTV      StreamingMovies  Contract PaperlessBilling  \\\n",
       "488                   Yes                   No  Two year              Yes   \n",
       "753   No internet service  No internet service  Two year               No   \n",
       "936                   Yes                  Yes  Two year               No   \n",
       "1082  No internet service  No internet service  Two year               No   \n",
       "1340                  Yes                   No  Two year               No   \n",
       "3331  No internet service  No internet service  Two year               No   \n",
       "3826  No internet service  No internet service  Two year               No   \n",
       "4380  No internet service  No internet service  Two year               No   \n",
       "5218  No internet service  No internet service  One year              Yes   \n",
       "6670                  Yes                   No  Two year               No   \n",
       "6754                   No                   No  Two year              Yes   \n",
       "\n",
       "                  PaymentMethod  MonthlyCharges TotalCharges Churn  \n",
       "488   Bank transfer (automatic)           52.55                 No  \n",
       "753                Mailed check           20.25                 No  \n",
       "936                Mailed check           80.85                 No  \n",
       "1082               Mailed check           25.75                 No  \n",
       "1340    Credit card (automatic)           56.05                 No  \n",
       "3331               Mailed check           19.85                 No  \n",
       "3826               Mailed check           25.35                 No  \n",
       "4380               Mailed check           20.00                 No  \n",
       "5218               Mailed check           19.70                 No  \n",
       "6670               Mailed check           73.35                 No  \n",
       "6754  Bank transfer (automatic)           61.90                 No  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[pd.to_numeric(df.TotalCharges,errors='coerce').isnull()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "86d39a8e-26a2-429c-a98c-0e9a51828c1f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(7043, 20)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "6f8ec96a-6619-4b11-ae64-f82317599c67",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "' '"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.iloc[488].TotalCharges"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "4d10e615-5278-4779-9af8-5c796b4ceb6b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(7032, 20)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.TotalCharges!=' '].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "57c83de4-30fc-4eaf-ad5d-fcb3aba85a2a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(7032, 20)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1 = df[df.TotalCharges!=' ']\n",
    "df1.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "0381e640-4345-4d99-afe5-3b54f2eacc6f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "gender               object\n",
       "SeniorCitizen         int64\n",
       "Partner              object\n",
       "Dependents           object\n",
       "tenure                int64\n",
       "PhoneService         object\n",
       "MultipleLines        object\n",
       "InternetService      object\n",
       "OnlineSecurity       object\n",
       "OnlineBackup         object\n",
       "DeviceProtection     object\n",
       "TechSupport          object\n",
       "StreamingTV          object\n",
       "StreamingMovies      object\n",
       "Contract             object\n",
       "PaperlessBilling     object\n",
       "PaymentMethod        object\n",
       "MonthlyCharges      float64\n",
       "TotalCharges         object\n",
       "Churn                object\n",
       "dtype: object"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "45a86732-eaf3-4c2b-956c-5eaf2c5209dc",
   "metadata": {},
   "outputs": [],
   "source": [
    "df1.TotalCharges = pd.to_numeric(df1.TotalCharges)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "5a7e914c-d357-44f6-9937-f27b1dd05544",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  29.85, 1889.5 ,  108.15, ...,  346.45,  306.6 , 6844.5 ])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.TotalCharges.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "2b6122bd-0aa7-41b2-9d57-e8a71aedf95b",
   "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>gender</th>\n",
       "      <th>SeniorCitizen</th>\n",
       "      <th>Partner</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>tenure</th>\n",
       "      <th>PhoneService</th>\n",
       "      <th>MultipleLines</th>\n",
       "      <th>InternetService</th>\n",
       "      <th>OnlineSecurity</th>\n",
       "      <th>OnlineBackup</th>\n",
       "      <th>DeviceProtection</th>\n",
       "      <th>TechSupport</th>\n",
       "      <th>StreamingTV</th>\n",
       "      <th>StreamingMovies</th>\n",
       "      <th>Contract</th>\n",
       "      <th>PaperlessBilling</th>\n",
       "      <th>PaymentMethod</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>No</td>\n",
       "      <td>No phone service</td>\n",
       "      <td>DSL</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>29.85</td>\n",
       "      <td>29.85</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>34</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>One year</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>56.95</td>\n",
       "      <td>1889.50</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>45</td>\n",
       "      <td>No</td>\n",
       "      <td>No phone service</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>One year</td>\n",
       "      <td>No</td>\n",
       "      <td>Bank transfer (automatic)</td>\n",
       "      <td>42.30</td>\n",
       "      <td>1840.75</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>22</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Fiber optic</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Credit card (automatic)</td>\n",
       "      <td>89.10</td>\n",
       "      <td>1949.40</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>10</td>\n",
       "      <td>No</td>\n",
       "      <td>No phone service</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>29.75</td>\n",
       "      <td>301.90</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7037</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>72</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>Two year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Bank transfer (automatic)</td>\n",
       "      <td>21.15</td>\n",
       "      <td>1419.40</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7038</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>24</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>One year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>84.80</td>\n",
       "      <td>1990.50</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7039</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>72</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Fiber optic</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>One year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Credit card (automatic)</td>\n",
       "      <td>103.20</td>\n",
       "      <td>7362.90</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7040</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>11</td>\n",
       "      <td>No</td>\n",
       "      <td>No phone service</td>\n",
       "      <td>DSL</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>29.60</td>\n",
       "      <td>346.45</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7042</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>66</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Fiber optic</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Two year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Bank transfer (automatic)</td>\n",
       "      <td>105.65</td>\n",
       "      <td>6844.50</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5163 rows × 20 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      gender  SeniorCitizen Partner Dependents  tenure PhoneService  \\\n",
       "0     Female              0     Yes         No       1           No   \n",
       "1       Male              0      No         No      34          Yes   \n",
       "3       Male              0      No         No      45           No   \n",
       "6       Male              0      No        Yes      22          Yes   \n",
       "7     Female              0      No         No      10           No   \n",
       "...      ...            ...     ...        ...     ...          ...   \n",
       "7037  Female              0      No         No      72          Yes   \n",
       "7038    Male              0     Yes        Yes      24          Yes   \n",
       "7039  Female              0     Yes        Yes      72          Yes   \n",
       "7040  Female              0     Yes        Yes      11           No   \n",
       "7042    Male              0      No         No      66          Yes   \n",
       "\n",
       "         MultipleLines InternetService       OnlineSecurity  \\\n",
       "0     No phone service             DSL                   No   \n",
       "1                   No             DSL                  Yes   \n",
       "3     No phone service             DSL                  Yes   \n",
       "6                  Yes     Fiber optic                   No   \n",
       "7     No phone service             DSL                  Yes   \n",
       "...                ...             ...                  ...   \n",
       "7037                No              No  No internet service   \n",
       "7038               Yes             DSL                  Yes   \n",
       "7039               Yes     Fiber optic                   No   \n",
       "7040  No phone service             DSL                  Yes   \n",
       "7042                No     Fiber optic                  Yes   \n",
       "\n",
       "             OnlineBackup     DeviceProtection          TechSupport  \\\n",
       "0                     Yes                   No                   No   \n",
       "1                      No                  Yes                   No   \n",
       "3                      No                  Yes                  Yes   \n",
       "6                     Yes                   No                   No   \n",
       "7                      No                   No                   No   \n",
       "...                   ...                  ...                  ...   \n",
       "7037  No internet service  No internet service  No internet service   \n",
       "7038                   No                  Yes                  Yes   \n",
       "7039                  Yes                  Yes                   No   \n",
       "7040                   No                   No                   No   \n",
       "7042                   No                  Yes                  Yes   \n",
       "\n",
       "              StreamingTV      StreamingMovies        Contract  \\\n",
       "0                      No                   No  Month-to-month   \n",
       "1                      No                   No        One year   \n",
       "3                      No                   No        One year   \n",
       "6                     Yes                   No  Month-to-month   \n",
       "7                      No                   No  Month-to-month   \n",
       "...                   ...                  ...             ...   \n",
       "7037  No internet service  No internet service        Two year   \n",
       "7038                  Yes                  Yes        One year   \n",
       "7039                  Yes                  Yes        One year   \n",
       "7040                   No                   No  Month-to-month   \n",
       "7042                  Yes                  Yes        Two year   \n",
       "\n",
       "     PaperlessBilling              PaymentMethod  MonthlyCharges  \\\n",
       "0                 Yes           Electronic check           29.85   \n",
       "1                  No               Mailed check           56.95   \n",
       "3                  No  Bank transfer (automatic)           42.30   \n",
       "6                 Yes    Credit card (automatic)           89.10   \n",
       "7                  No               Mailed check           29.75   \n",
       "...               ...                        ...             ...   \n",
       "7037              Yes  Bank transfer (automatic)           21.15   \n",
       "7038              Yes               Mailed check           84.80   \n",
       "7039              Yes    Credit card (automatic)          103.20   \n",
       "7040              Yes           Electronic check           29.60   \n",
       "7042              Yes  Bank transfer (automatic)          105.65   \n",
       "\n",
       "      TotalCharges Churn  \n",
       "0            29.85    No  \n",
       "1          1889.50    No  \n",
       "3          1840.75    No  \n",
       "6          1949.40    No  \n",
       "7           301.90    No  \n",
       "...            ...   ...  \n",
       "7037       1419.40    No  \n",
       "7038       1990.50    No  \n",
       "7039       7362.90    No  \n",
       "7040        346.45    No  \n",
       "7042       6844.50    No  \n",
       "\n",
       "[5163 rows x 20 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1[df1.Churn=='No']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e068b68-45e3-4b07-ba93-089ee678a278",
   "metadata": {},
   "source": [
    "## Data Visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "b0f6b8c7-cfac-4679-91ca-9749a2fc4098",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x132e33070>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tenure_churn_no = df1[df1.Churn=='No'].tenure\n",
    "tenure_churn_yes = df1[df1.Churn=='Yes'].tenure\n",
    "\n",
    "plt.xlabel(\"tenure\")\n",
    "plt.ylabel(\"Number Of Customers\")\n",
    "plt.title(\"Customer Churn Prediction Visualiztion\")\n",
    "\n",
    "plt.hist([tenure_churn_yes, tenure_churn_no], rwidth=0.95, color=['green','red'],label=['Churn=Yes','Churn=No'])\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "c787d680-53db-4742-845c-d9f267b1b610",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x13305ddf0>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mc_churn_no = df1[df1.Churn=='No'].MonthlyCharges      \n",
    "mc_churn_yes = df1[df1.Churn=='Yes'].MonthlyCharges      \n",
    "\n",
    "plt.xlabel(\"Monthly Charges\")\n",
    "plt.ylabel(\"Number Of Customers\")\n",
    "plt.title(\"Customer Churn Prediction Visualiztion\")\n",
    "\n",
    "plt.hist([mc_churn_yes, mc_churn_no], rwidth=0.95, color=['green','red'],label=['Churn=Yes','Churn=No'])\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "913efb36-7224-4098-b031-204cb4bd713a",
   "metadata": {},
   "outputs": [],
   "source": [
    "def print_unique_col_values(df):\n",
    "       for column in df:\n",
    "            if df[column].dtypes=='object':\n",
    "                print(f'{column}: {df[column].unique()}') \n",
    "                "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "7ba732e7-3e08-4572-9610-228048f32f7d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gender: ['Female' 'Male']\n",
      "Partner: ['Yes' 'No']\n",
      "Dependents: ['No' 'Yes']\n",
      "PhoneService: ['No' 'Yes']\n",
      "MultipleLines: ['No phone service' 'No' 'Yes']\n",
      "InternetService: ['DSL' 'Fiber optic' 'No']\n",
      "OnlineSecurity: ['No' 'Yes' 'No internet service']\n",
      "OnlineBackup: ['Yes' 'No' 'No internet service']\n",
      "DeviceProtection: ['No' 'Yes' 'No internet service']\n",
      "TechSupport: ['No' 'Yes' 'No internet service']\n",
      "StreamingTV: ['No' 'Yes' 'No internet service']\n",
      "StreamingMovies: ['No' 'Yes' 'No internet service']\n",
      "Contract: ['Month-to-month' 'One year' 'Two year']\n",
      "PaperlessBilling: ['Yes' 'No']\n",
      "PaymentMethod: ['Electronic check' 'Mailed check' 'Bank transfer (automatic)'\n",
      " 'Credit card (automatic)']\n",
      "Churn: ['No' 'Yes']\n"
     ]
    }
   ],
   "source": [
    "print_unique_col_values(df1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "3787fa6c-bdfb-4d61-80a8-84509564c0e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "df1.replace('No internet service','No',inplace=True)\n",
    "df1.replace('No phone service','No',inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "3b71a78e-5466-48cb-8cfd-06731ecf3d14",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gender: ['Female' 'Male']\n",
      "Partner: ['Yes' 'No']\n",
      "Dependents: ['No' 'Yes']\n",
      "PhoneService: ['No' 'Yes']\n",
      "MultipleLines: ['No' 'Yes']\n",
      "InternetService: ['DSL' 'Fiber optic' 'No']\n",
      "OnlineSecurity: ['No' 'Yes']\n",
      "OnlineBackup: ['Yes' 'No']\n",
      "DeviceProtection: ['No' 'Yes']\n",
      "TechSupport: ['No' 'Yes']\n",
      "StreamingTV: ['No' 'Yes']\n",
      "StreamingMovies: ['No' 'Yes']\n",
      "Contract: ['Month-to-month' 'One year' 'Two year']\n",
      "PaperlessBilling: ['Yes' 'No']\n",
      "PaymentMethod: ['Electronic check' 'Mailed check' 'Bank transfer (automatic)'\n",
      " 'Credit card (automatic)']\n",
      "Churn: ['No' 'Yes']\n"
     ]
    }
   ],
   "source": [
    "print_unique_col_values(df1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "32d2c1fa-f6ee-4dfd-b683-cce73c4ea396",
   "metadata": {},
   "outputs": [],
   "source": [
    "yes_no_columns = ['Partner','Dependents','PhoneService','MultipleLines','OnlineSecurity','OnlineBackup',\n",
    "                  'DeviceProtection','TechSupport','StreamingTV','StreamingMovies','PaperlessBilling','Churn']\n",
    "for col in yes_no_columns:\n",
    "    df1[col].replace({'Yes': 1,'No': 0},inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "234195cc-2de9-4c6c-8809-ff39ab086bff",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gender: ['Female' 'Male']\n",
      "SeniorCitizen: [0 1]\n",
      "Partner: [1 0]\n",
      "Dependents: [0 1]\n",
      "tenure: [ 1 34  2 45  8 22 10 28 62 13 16 58 49 25 69 52 71 21 12 30 47 72 17 27\n",
      "  5 46 11 70 63 43 15 60 18 66  9  3 31 50 64 56  7 42 35 48 29 65 38 68\n",
      " 32 55 37 36 41  6  4 33 67 23 57 61 14 20 53 40 59 24 44 19 54 51 26 39]\n",
      "PhoneService: [0 1]\n",
      "MultipleLines: [0 1]\n",
      "InternetService: ['DSL' 'Fiber optic' 'No']\n",
      "OnlineSecurity: [0 1]\n",
      "OnlineBackup: [1 0]\n",
      "DeviceProtection: [0 1]\n",
      "TechSupport: [0 1]\n",
      "StreamingTV: [0 1]\n",
      "StreamingMovies: [0 1]\n",
      "Contract: ['Month-to-month' 'One year' 'Two year']\n",
      "PaperlessBilling: [1 0]\n",
      "PaymentMethod: ['Electronic check' 'Mailed check' 'Bank transfer (automatic)'\n",
      " 'Credit card (automatic)']\n",
      "MonthlyCharges: [29.85 56.95 53.85 ... 63.1  44.2  78.7 ]\n",
      "TotalCharges: [  29.85 1889.5   108.15 ...  346.45  306.6  6844.5 ]\n",
      "Churn: [0 1]\n"
     ]
    }
   ],
   "source": [
    "for col in df1:\n",
    "    print(f'{col}: {df1[col].unique()}') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "a66c78d8-daa7-4f92-860f-41a293f3b5f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "df1['gender'].replace({'Female':1,'Male':0},inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "48c7fe1e-1874-419b-8a23-2011218bd714",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 0])"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.gender.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "56fdfb23-cdff-455d-b760-60ae740bfcb5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['gender', 'SeniorCitizen', 'Partner', 'Dependents', 'tenure',\n",
       "       'PhoneService', 'MultipleLines', 'OnlineSecurity', 'OnlineBackup',\n",
       "       'DeviceProtection', 'TechSupport', 'StreamingTV', 'StreamingMovies',\n",
       "       'PaperlessBilling', 'MonthlyCharges', 'TotalCharges', 'Churn',\n",
       "       'InternetService_DSL', 'InternetService_Fiber optic',\n",
       "       'InternetService_No', 'Contract_Month-to-month', 'Contract_One year',\n",
       "       'Contract_Two year', 'PaymentMethod_Bank transfer (automatic)',\n",
       "       'PaymentMethod_Credit card (automatic)',\n",
       "       'PaymentMethod_Electronic check', 'PaymentMethod_Mailed check'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2 = pd.get_dummies(data=df1, columns=['InternetService','Contract','PaymentMethod'])\n",
    "df2.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "9efd6b3e-5f21-484b-b6cf-81dcf7ad3d62",
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_scale = ['tenure','MonthlyCharges','TotalCharges']\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "scaler = MinMaxScaler()\n",
    "df2[cols_to_scale] = scaler.fit_transform(df2[cols_to_scale])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "a4476a88-065d-4c6a-954b-e1065ddbc6e5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gender: [1 0]\n",
      "SeniorCitizen: [0 1]\n",
      "Partner: [1 0]\n",
      "Dependents: [0 1]\n",
      "tenure: [0.         0.46478873 0.01408451 0.61971831 0.09859155 0.29577465\n",
      " 0.12676056 0.38028169 0.85915493 0.16901408 0.21126761 0.8028169\n",
      " 0.67605634 0.33802817 0.95774648 0.71830986 0.98591549 0.28169014\n",
      " 0.15492958 0.4084507  0.64788732 1.         0.22535211 0.36619718\n",
      " 0.05633803 0.63380282 0.14084507 0.97183099 0.87323944 0.5915493\n",
      " 0.1971831  0.83098592 0.23943662 0.91549296 0.11267606 0.02816901\n",
      " 0.42253521 0.69014085 0.88732394 0.77464789 0.08450704 0.57746479\n",
      " 0.47887324 0.66197183 0.3943662  0.90140845 0.52112676 0.94366197\n",
      " 0.43661972 0.76056338 0.50704225 0.49295775 0.56338028 0.07042254\n",
      " 0.04225352 0.45070423 0.92957746 0.30985915 0.78873239 0.84507042\n",
      " 0.18309859 0.26760563 0.73239437 0.54929577 0.81690141 0.32394366\n",
      " 0.6056338  0.25352113 0.74647887 0.70422535 0.35211268 0.53521127]\n",
      "PhoneService: [0 1]\n",
      "MultipleLines: [0 1]\n",
      "OnlineSecurity: [0 1]\n",
      "OnlineBackup: [1 0]\n",
      "DeviceProtection: [0 1]\n",
      "TechSupport: [0 1]\n",
      "StreamingTV: [0 1]\n",
      "StreamingMovies: [0 1]\n",
      "PaperlessBilling: [1 0]\n",
      "MonthlyCharges: [0.11542289 0.38507463 0.35422886 ... 0.44626866 0.25820896 0.60149254]\n",
      "TotalCharges: [0.0012751  0.21586661 0.01031041 ... 0.03780868 0.03321025 0.78764136]\n",
      "Churn: [0 1]\n",
      "InternetService_DSL: [1 0]\n",
      "InternetService_Fiber optic: [0 1]\n",
      "InternetService_No: [0 1]\n",
      "Contract_Month-to-month: [1 0]\n",
      "Contract_One year: [0 1]\n",
      "Contract_Two year: [0 1]\n",
      "PaymentMethod_Bank transfer (automatic): [0 1]\n",
      "PaymentMethod_Credit card (automatic): [0 1]\n",
      "PaymentMethod_Electronic check: [1 0]\n",
      "PaymentMethod_Mailed check: [0 1]\n"
     ]
    }
   ],
   "source": [
    "for col in df2:\n",
    "    print(f'{col}: {df2[col].unique()}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca43f120-1cf8-4ec0-8f41-1f3f3ae1d0b9",
   "metadata": {},
   "source": [
    "## Train Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "5c7eb4d0-e762-469b-97b6-342e225b2d3e",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df2.drop('Churn',axis='columns')\n",
    "y = testLabels = df2.Churn.astype(np.float32)\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=15, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "5bcfaff1-8b47-49c2-89c6-00b6d0a6392f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0    4130\n",
       "1.0    1495\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "37d19c69-7cf6-4357-9394-89356eb8db29",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0    5163\n",
       "1.0    1869\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "d9618ce8-60f1-437b-815c-3bdd1f33e16f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0    1033\n",
       "1.0     374\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "03ff6913-01a1-4261-b963-649772ad5b7f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5625, 26)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "f993005c-2a64-4407-bd49-b592c014f2cc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1407, 26)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "79609644-24b7-433e-a978-a478f114c1bc",
   "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>gender</th>\n",
       "      <th>SeniorCitizen</th>\n",
       "      <th>Partner</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>tenure</th>\n",
       "      <th>PhoneService</th>\n",
       "      <th>MultipleLines</th>\n",
       "      <th>OnlineSecurity</th>\n",
       "      <th>OnlineBackup</th>\n",
       "      <th>DeviceProtection</th>\n",
       "      <th>...</th>\n",
       "      <th>InternetService_DSL</th>\n",
       "      <th>InternetService_Fiber optic</th>\n",
       "      <th>InternetService_No</th>\n",
       "      <th>Contract_Month-to-month</th>\n",
       "      <th>Contract_One year</th>\n",
       "      <th>Contract_Two year</th>\n",
       "      <th>PaymentMethod_Bank transfer (automatic)</th>\n",
       "      <th>PaymentMethod_Credit card (automatic)</th>\n",
       "      <th>PaymentMethod_Electronic check</th>\n",
       "      <th>PaymentMethod_Mailed check</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>684</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2446</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.239437</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1680</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.774648</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2220</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2842</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.042254</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5131</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.605634</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4372</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4687</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.295775</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2385</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.028169</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6773</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.507042</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      gender  SeniorCitizen  Partner  Dependents    tenure  PhoneService  \\\n",
       "684        1              0        0           0  0.000000             1   \n",
       "2446       1              0        0           0  0.239437             1   \n",
       "1680       0              0        1           1  0.774648             1   \n",
       "2220       0              0        1           0  1.000000             1   \n",
       "2842       1              0        0           0  0.042254             0   \n",
       "5131       1              0        1           0  0.605634             1   \n",
       "4372       0              0        1           1  0.000000             1   \n",
       "4687       0              0        1           0  0.295775             1   \n",
       "2385       1              0        0           1  0.028169             0   \n",
       "6773       0              1        1           0  0.507042             1   \n",
       "\n",
       "      MultipleLines  OnlineSecurity  OnlineBackup  DeviceProtection  ...  \\\n",
       "684               0               0             0                 0  ...   \n",
       "2446              1               0             1                 0  ...   \n",
       "1680              1               0             0                 0  ...   \n",
       "2220              0               1             1                 0  ...   \n",
       "2842              0               1             0                 1  ...   \n",
       "5131              1               0             0                 0  ...   \n",
       "4372              0               0             0                 0  ...   \n",
       "4687              0               0             0                 1  ...   \n",
       "2385              0               0             1                 0  ...   \n",
       "6773              1               0             0                 0  ...   \n",
       "\n",
       "      InternetService_DSL  InternetService_Fiber optic  InternetService_No  \\\n",
       "684                     0                            1                   0   \n",
       "2446                    0                            1                   0   \n",
       "1680                    0                            0                   1   \n",
       "2220                    1                            0                   0   \n",
       "2842                    1                            0                   0   \n",
       "5131                    0                            1                   0   \n",
       "4372                    0                            0                   1   \n",
       "4687                    0                            1                   0   \n",
       "2385                    1                            0                   0   \n",
       "6773                    0                            1                   0   \n",
       "\n",
       "      Contract_Month-to-month  Contract_One year  Contract_Two year  \\\n",
       "684                         1                  0                  0   \n",
       "2446                        1                  0                  0   \n",
       "1680                        0                  1                  0   \n",
       "2220                        0                  0                  1   \n",
       "2842                        1                  0                  0   \n",
       "5131                        1                  0                  0   \n",
       "4372                        1                  0                  0   \n",
       "4687                        1                  0                  0   \n",
       "2385                        1                  0                  0   \n",
       "6773                        1                  0                  0   \n",
       "\n",
       "      PaymentMethod_Bank transfer (automatic)  \\\n",
       "684                                         0   \n",
       "2446                                        0   \n",
       "1680                                        0   \n",
       "2220                                        1   \n",
       "2842                                        0   \n",
       "5131                                        0   \n",
       "4372                                        1   \n",
       "4687                                        0   \n",
       "2385                                        0   \n",
       "6773                                        0   \n",
       "\n",
       "      PaymentMethod_Credit card (automatic)  PaymentMethod_Electronic check  \\\n",
       "684                                       0                               0   \n",
       "2446                                      1                               0   \n",
       "1680                                      0                               0   \n",
       "2220                                      0                               0   \n",
       "2842                                      0                               0   \n",
       "5131                                      0                               1   \n",
       "4372                                      0                               0   \n",
       "4687                                      1                               0   \n",
       "2385                                      0                               0   \n",
       "6773                                      0                               1   \n",
       "\n",
       "      PaymentMethod_Mailed check  \n",
       "684                            1  \n",
       "2446                           0  \n",
       "1680                           1  \n",
       "2220                           0  \n",
       "2842                           1  \n",
       "5131                           0  \n",
       "4372                           0  \n",
       "4687                           0  \n",
       "2385                           1  \n",
       "6773                           0  \n",
       "\n",
       "[10 rows x 26 columns]"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "9a8d0653-72eb-4c7d-93d1-71944caee94d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "26"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(X_train.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "435918a2-8747-4967-9e42-67c6b9fece07",
   "metadata": {},
   "source": [
    "## Build a model (ANN) in tensorflow/keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "8464b4b6-cf5d-4d4b-b0e1-bd2e1239dac3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting tensorflow_addons\n",
      "  Downloading tensorflow_addons-0.19.0-cp39-cp39-macosx_11_0_arm64.whl (11.1 MB)\n",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m11.1/11.1 MB\u001b[0m \u001b[31m4.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n",
      "\u001b[?25hRequirement already satisfied: packaging in /Users/serhatderya/miniconda3/envs/tfm1_arm64/lib/python3.9/site-packages (from tensorflow_addons) (23.0)\n",
      "Collecting typeguard>=2.7\n",
      "  Downloading typeguard-2.13.3-py3-none-any.whl (17 kB)\n",
      "Installing collected packages: typeguard, tensorflow_addons\n",
      "Successfully installed tensorflow_addons-0.19.0 typeguard-2.13.3\n"
     ]
    }
   ],
   "source": [
    "!pip install tensorflow_addons"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "8769c480-1b6a-4bbe-b074-6e3a6b2a1bb6",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow_addons import losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "d4d4a464-cba6-443f-8985-d2e3e2fedfed",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from sklearn.metrics import confusion_matrix , classification_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "fd48c8cd-71c3-411f-8ceb-497e395c0508",
   "metadata": {},
   "outputs": [],
   "source": [
    "def ANN(X_train, y_train, X_test, y_test, loss, weights):\n",
    "    model = keras.Sequential([\n",
    "        keras.layers.Dense(26, input_dim=26, activation='relu'),\n",
    "        keras.layers.Dense(15, activation='relu'),\n",
    "        keras.layers.Dense(1, activation='sigmoid')\n",
    "    ])\n",
    "\n",
    "    model.compile(optimizer='adam', loss=loss, metrics=['accuracy'])\n",
    "    \n",
    "    if weights == -1:\n",
    "        model.fit(X_train, y_train, epochs=100)\n",
    "    else:\n",
    "        model.fit(X_train, y_train, epochs=100, class_weight = weights)\n",
    "    \n",
    "    print(model.evaluate(X_test, y_test))\n",
    "    \n",
    "    y_preds = model.predict(X_test)\n",
    "    y_preds = np.round(y_preds)\n",
    "    \n",
    "    print(\"Classification Report: \\n\", classification_report(y_test, y_preds))\n",
    "    \n",
    "    return y_preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "5e792ac1-e240-41d4-80c9-124d7b7fb9d0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Metal device set to: Apple M1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:12:58.192570: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:305] Could not identify NUMA node of platform GPU ID 0, defaulting to 0. Your kernel may not have been built with NUMA support.\n",
      "2023-01-20 09:12:58.192837: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:271] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 0 MB memory) -> physical PluggableDevice (device: 0, name: METAL, pci bus id: <undefined>)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:12:58.695390: W tensorflow/core/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n",
      "2023-01-20 09:12:58.861878: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "176/176 [==============================] - 2s 6ms/step - loss: 0.4895 - accuracy: 0.7570\n",
      "Epoch 2/100\n",
      "176/176 [==============================] - 1s 5ms/step - loss: 0.4289 - accuracy: 0.7888\n",
      "Epoch 3/100\n",
      "176/176 [==============================] - 1s 5ms/step - loss: 0.4221 - accuracy: 0.7956\n",
      "Epoch 4/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4186 - accuracy: 0.8000\n",
      "Epoch 5/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4162 - accuracy: 0.8007\n",
      "Epoch 6/100\n",
      "176/176 [==============================] - 1s 5ms/step - loss: 0.4148 - accuracy: 0.8032\n",
      "Epoch 7/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4137 - accuracy: 0.8021\n",
      "Epoch 8/100\n",
      "176/176 [==============================] - 1s 5ms/step - loss: 0.4123 - accuracy: 0.8027\n",
      "Epoch 9/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4107 - accuracy: 0.8050\n",
      "Epoch 10/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4100 - accuracy: 0.8039\n",
      "Epoch 11/100\n",
      "176/176 [==============================] - 1s 5ms/step - loss: 0.4083 - accuracy: 0.8055\n",
      "Epoch 12/100\n",
      "176/176 [==============================] - 1s 5ms/step - loss: 0.4078 - accuracy: 0.8055\n",
      "Epoch 13/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4058 - accuracy: 0.8075\n",
      "Epoch 14/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4052 - accuracy: 0.8069\n",
      "Epoch 15/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4042 - accuracy: 0.8069\n",
      "Epoch 16/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4036 - accuracy: 0.8066\n",
      "Epoch 17/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4037 - accuracy: 0.8092\n",
      "Epoch 18/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4020 - accuracy: 0.8073\n",
      "Epoch 19/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.4014 - accuracy: 0.8084\n",
      "Epoch 20/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3998 - accuracy: 0.8123\n",
      "Epoch 21/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3991 - accuracy: 0.8082\n",
      "Epoch 22/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3965 - accuracy: 0.8105\n",
      "Epoch 23/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3964 - accuracy: 0.8091\n",
      "Epoch 24/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3952 - accuracy: 0.8130\n",
      "Epoch 25/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3944 - accuracy: 0.8128\n",
      "Epoch 26/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3925 - accuracy: 0.8114\n",
      "Epoch 27/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3920 - accuracy: 0.8148\n",
      "Epoch 28/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3911 - accuracy: 0.8092\n",
      "Epoch 29/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3908 - accuracy: 0.8133\n",
      "Epoch 30/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3885 - accuracy: 0.8119\n",
      "Epoch 31/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3884 - accuracy: 0.8117\n",
      "Epoch 32/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3878 - accuracy: 0.8121\n",
      "Epoch 33/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3878 - accuracy: 0.8091\n",
      "Epoch 34/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3854 - accuracy: 0.8135\n",
      "Epoch 35/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3841 - accuracy: 0.8160\n",
      "Epoch 36/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3829 - accuracy: 0.8160\n",
      "Epoch 37/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3823 - accuracy: 0.8117\n",
      "Epoch 38/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3820 - accuracy: 0.8149\n",
      "Epoch 39/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3805 - accuracy: 0.8148\n",
      "Epoch 40/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3811 - accuracy: 0.8133\n",
      "Epoch 41/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3789 - accuracy: 0.8140\n",
      "Epoch 42/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3786 - accuracy: 0.8151\n",
      "Epoch 43/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3778 - accuracy: 0.8156\n",
      "Epoch 44/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3760 - accuracy: 0.8176\n",
      "Epoch 45/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3763 - accuracy: 0.8185\n",
      "Epoch 46/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3742 - accuracy: 0.8178\n",
      "Epoch 47/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3740 - accuracy: 0.8199\n",
      "Epoch 48/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3730 - accuracy: 0.8203\n",
      "Epoch 49/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3713 - accuracy: 0.8199\n",
      "Epoch 50/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3724 - accuracy: 0.8164\n",
      "Epoch 51/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3701 - accuracy: 0.8213\n",
      "Epoch 52/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3703 - accuracy: 0.8226\n",
      "Epoch 53/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3696 - accuracy: 0.8213\n",
      "Epoch 54/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3697 - accuracy: 0.8219\n",
      "Epoch 55/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3679 - accuracy: 0.8190\n",
      "Epoch 56/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3671 - accuracy: 0.8247\n",
      "Epoch 57/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3664 - accuracy: 0.8258\n",
      "Epoch 58/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3665 - accuracy: 0.8251\n",
      "Epoch 59/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3653 - accuracy: 0.8251\n",
      "Epoch 60/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3641 - accuracy: 0.8267\n",
      "Epoch 61/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3646 - accuracy: 0.8231\n",
      "Epoch 62/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3647 - accuracy: 0.8244\n",
      "Epoch 63/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3644 - accuracy: 0.8233\n",
      "Epoch 64/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3631 - accuracy: 0.8274\n",
      "Epoch 65/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3626 - accuracy: 0.8267\n",
      "Epoch 66/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3615 - accuracy: 0.8272\n",
      "Epoch 67/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3616 - accuracy: 0.8277\n",
      "Epoch 68/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3613 - accuracy: 0.8277\n",
      "Epoch 69/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3597 - accuracy: 0.8293\n",
      "Epoch 70/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3594 - accuracy: 0.8274\n",
      "Epoch 71/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3595 - accuracy: 0.8286\n",
      "Epoch 72/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3582 - accuracy: 0.8292\n",
      "Epoch 73/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3580 - accuracy: 0.8256\n",
      "Epoch 74/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3580 - accuracy: 0.8284\n",
      "Epoch 75/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3588 - accuracy: 0.8256\n",
      "Epoch 76/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3552 - accuracy: 0.8293\n",
      "Epoch 77/100\n",
      "176/176 [==============================] - 1s 5ms/step - loss: 0.3560 - accuracy: 0.8300\n",
      "Epoch 78/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3553 - accuracy: 0.8327\n",
      "Epoch 79/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3554 - accuracy: 0.8315\n",
      "Epoch 80/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3552 - accuracy: 0.8320\n",
      "Epoch 81/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3538 - accuracy: 0.8327\n",
      "Epoch 82/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3536 - accuracy: 0.8329\n",
      "Epoch 83/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3545 - accuracy: 0.8299\n",
      "Epoch 84/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3546 - accuracy: 0.8302\n",
      "Epoch 85/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3530 - accuracy: 0.8309\n",
      "Epoch 86/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3512 - accuracy: 0.8306\n",
      "Epoch 87/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3521 - accuracy: 0.8318\n",
      "Epoch 88/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3501 - accuracy: 0.8336\n",
      "Epoch 89/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3514 - accuracy: 0.8322\n",
      "Epoch 90/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3501 - accuracy: 0.8352\n",
      "Epoch 91/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3499 - accuracy: 0.8322\n",
      "Epoch 92/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3512 - accuracy: 0.8322\n",
      "Epoch 93/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3506 - accuracy: 0.8348\n",
      "Epoch 94/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3491 - accuracy: 0.8320\n",
      "Epoch 95/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3491 - accuracy: 0.8372\n",
      "Epoch 96/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3485 - accuracy: 0.8341\n",
      "Epoch 97/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3469 - accuracy: 0.8359\n",
      "Epoch 98/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3474 - accuracy: 0.8341\n",
      "Epoch 99/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3464 - accuracy: 0.8347\n",
      "Epoch 100/100\n",
      "176/176 [==============================] - 1s 6ms/step - loss: 0.3459 - accuracy: 0.8354\n",
      "28/44 [==================>...........] - ETA: 0s - loss: 0.4759 - accuracy: 0.7946"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:14:40.332388: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44/44 [==============================] - 0s 5ms/step - loss: 0.4857 - accuracy: 0.7854\n",
      "[0.48569804430007935, 0.7853589653968811]\n",
      "44/44 [==============================] - 0s 2ms/step\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.82      0.91      0.86      1033\n",
      "         1.0       0.64      0.45      0.53       374\n",
      "\n",
      "    accuracy                           0.79      1407\n",
      "   macro avg       0.73      0.68      0.69      1407\n",
      "weighted avg       0.77      0.79      0.77      1407\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:14:40.626240: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    }
   ],
   "source": [
    "y_preds = ANN(X_train, y_train, X_test, y_test, 'binary_crossentropy', -1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fdca42b8-f59d-4c40-b99e-d142e08d6bba",
   "metadata": {},
   "source": [
    "# Mitigating Skewdness of Data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1be0087d-8cbe-46ea-bf67-969b712b83f2",
   "metadata": {},
   "source": [
    "## Method 1: Undersampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "f6c24de9-bcc5-48d9-8fce-3e1c4612eb87",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Class Count\n",
    "count_class_0, count_class_1 = df1.Churn.value_counts()\n",
    "\n",
    "# Divide by class\n",
    "df_class_0 = df2[df2[\"Churn\"] == 0]\n",
    "df_class_1 = df2[df2[\"Churn\"] == 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "037c5828-83b0-4c64-bdfa-278f4e6f17ba",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random under-sampling:\n",
      "0    1869\n",
      "1    1869\n",
      "Name: Churn, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# Undersample 0-class and concat the DataFrames of both classes\n",
    "df_class_0_under = df_class_0.sample(count_class_1)\n",
    "df_test_under = pd.concat([df_class_0_under, df_class_1], axis=0)\n",
    "\n",
    "print(\"Random under-sampling:\")\n",
    "print(df_test_under.Churn.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "3aa45bef-2f32-4509-b42a-1083193d60c7",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df_test_under.drop(\"Churn\", axis=\"columns\")\n",
    "y = df_test_under[\"Churn\"]\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=15, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "d7e6aa3e-8e1c-49fd-a000-b3fda3f49c61",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1495\n",
       "1    1495\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of classes in training Data\n",
    "y_train.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "b14e6a9f-267e-42af-8a16-78f21f9032a9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      " 9/94 [=>............................] - ETA: 0s - loss: 0.6953 - accuracy: 0.5139 "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:23:12.174189: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "94/94 [==============================] - 1s 9ms/step - loss: 0.6442 - accuracy: 0.6415\n",
      "Epoch 2/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.5324 - accuracy: 0.7478\n",
      "Epoch 3/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.5033 - accuracy: 0.7589\n",
      "Epoch 4/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4935 - accuracy: 0.7686\n",
      "Epoch 5/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4894 - accuracy: 0.7662\n",
      "Epoch 6/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4851 - accuracy: 0.7702\n",
      "Epoch 7/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4818 - accuracy: 0.7699\n",
      "Epoch 8/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4807 - accuracy: 0.7696\n",
      "Epoch 9/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4766 - accuracy: 0.7729\n",
      "Epoch 10/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4763 - accuracy: 0.7739\n",
      "Epoch 11/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4752 - accuracy: 0.7726\n",
      "Epoch 12/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4720 - accuracy: 0.7732\n",
      "Epoch 13/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4714 - accuracy: 0.7736\n",
      "Epoch 14/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4693 - accuracy: 0.7783\n",
      "Epoch 15/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4704 - accuracy: 0.7793\n",
      "Epoch 16/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4684 - accuracy: 0.7779\n",
      "Epoch 17/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4656 - accuracy: 0.7819\n",
      "Epoch 18/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4644 - accuracy: 0.7853\n",
      "Epoch 19/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4638 - accuracy: 0.7809\n",
      "Epoch 20/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4635 - accuracy: 0.7813\n",
      "Epoch 21/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4623 - accuracy: 0.7839\n",
      "Epoch 22/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4614 - accuracy: 0.7853\n",
      "Epoch 23/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4597 - accuracy: 0.7833\n",
      "Epoch 24/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4590 - accuracy: 0.7856\n",
      "Epoch 25/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4604 - accuracy: 0.7813\n",
      "Epoch 26/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4572 - accuracy: 0.7876\n",
      "Epoch 27/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4557 - accuracy: 0.7833\n",
      "Epoch 28/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4552 - accuracy: 0.7839\n",
      "Epoch 29/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4568 - accuracy: 0.7883\n",
      "Epoch 30/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4529 - accuracy: 0.7843\n",
      "Epoch 31/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4538 - accuracy: 0.7893\n",
      "Epoch 32/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4531 - accuracy: 0.7890\n",
      "Epoch 33/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4504 - accuracy: 0.7946\n",
      "Epoch 34/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4516 - accuracy: 0.7883\n",
      "Epoch 35/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4494 - accuracy: 0.7953\n",
      "Epoch 36/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4499 - accuracy: 0.7916\n",
      "Epoch 37/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4480 - accuracy: 0.7933\n",
      "Epoch 38/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4468 - accuracy: 0.7946\n",
      "Epoch 39/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4459 - accuracy: 0.7967\n",
      "Epoch 40/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4444 - accuracy: 0.7963\n",
      "Epoch 41/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4448 - accuracy: 0.7970\n",
      "Epoch 42/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4413 - accuracy: 0.7997\n",
      "Epoch 43/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4434 - accuracy: 0.7990\n",
      "Epoch 44/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4402 - accuracy: 0.8013\n",
      "Epoch 45/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4422 - accuracy: 0.8007\n",
      "Epoch 46/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4384 - accuracy: 0.8013\n",
      "Epoch 47/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4388 - accuracy: 0.8010\n",
      "Epoch 48/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4373 - accuracy: 0.8013\n",
      "Epoch 49/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4376 - accuracy: 0.8007\n",
      "Epoch 50/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4365 - accuracy: 0.7990\n",
      "Epoch 51/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4367 - accuracy: 0.7977\n",
      "Epoch 52/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4348 - accuracy: 0.8020\n",
      "Epoch 53/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4331 - accuracy: 0.8040\n",
      "Epoch 54/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4348 - accuracy: 0.7973\n",
      "Epoch 55/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4326 - accuracy: 0.8054\n",
      "Epoch 56/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4313 - accuracy: 0.8023\n",
      "Epoch 57/100\n",
      "94/94 [==============================] - 1s 7ms/step - loss: 0.4319 - accuracy: 0.7997\n",
      "Epoch 58/100\n",
      "94/94 [==============================] - 1s 7ms/step - loss: 0.4317 - accuracy: 0.7987\n",
      "Epoch 59/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4282 - accuracy: 0.8064\n",
      "Epoch 60/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4307 - accuracy: 0.8017\n",
      "Epoch 61/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4289 - accuracy: 0.8057\n",
      "Epoch 62/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4277 - accuracy: 0.8030\n",
      "Epoch 63/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4267 - accuracy: 0.8054\n",
      "Epoch 64/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4251 - accuracy: 0.8104\n",
      "Epoch 65/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4279 - accuracy: 0.8017\n",
      "Epoch 66/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4248 - accuracy: 0.8033\n",
      "Epoch 67/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4238 - accuracy: 0.8084\n",
      "Epoch 68/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4231 - accuracy: 0.8070\n",
      "Epoch 69/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4235 - accuracy: 0.8114\n",
      "Epoch 70/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4233 - accuracy: 0.8080\n",
      "Epoch 71/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4227 - accuracy: 0.8084\n",
      "Epoch 72/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4221 - accuracy: 0.8057\n",
      "Epoch 73/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4204 - accuracy: 0.8107\n",
      "Epoch 74/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4212 - accuracy: 0.8057\n",
      "Epoch 75/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4185 - accuracy: 0.8151\n",
      "Epoch 76/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4199 - accuracy: 0.8060\n",
      "Epoch 77/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4201 - accuracy: 0.8050\n",
      "Epoch 78/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4197 - accuracy: 0.8080\n",
      "Epoch 79/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4171 - accuracy: 0.8117\n",
      "Epoch 80/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4209 - accuracy: 0.8094\n",
      "Epoch 81/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4162 - accuracy: 0.8134\n",
      "Epoch 82/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4176 - accuracy: 0.8070\n",
      "Epoch 83/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4144 - accuracy: 0.8100\n",
      "Epoch 84/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4152 - accuracy: 0.8100\n",
      "Epoch 85/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4138 - accuracy: 0.8144\n",
      "Epoch 86/100\n",
      "94/94 [==============================] - 1s 7ms/step - loss: 0.4126 - accuracy: 0.8120\n",
      "Epoch 87/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4123 - accuracy: 0.8130\n",
      "Epoch 88/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4127 - accuracy: 0.8100\n",
      "Epoch 89/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4132 - accuracy: 0.8114\n",
      "Epoch 90/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4104 - accuracy: 0.8097\n",
      "Epoch 91/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4103 - accuracy: 0.8154\n",
      "Epoch 92/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4097 - accuracy: 0.8104\n",
      "Epoch 93/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4090 - accuracy: 0.8161\n",
      "Epoch 94/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4082 - accuracy: 0.8181\n",
      "Epoch 95/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4076 - accuracy: 0.8147\n",
      "Epoch 96/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4076 - accuracy: 0.8151\n",
      "Epoch 97/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4060 - accuracy: 0.8137\n",
      "Epoch 98/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4057 - accuracy: 0.8151\n",
      "Epoch 99/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4053 - accuracy: 0.8097\n",
      "Epoch 100/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4048 - accuracy: 0.8130\n",
      "14/24 [================>.............] - ETA: 0s - loss: 0.5679 - accuracy: 0.7210"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:24:08.048404: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "24/24 [==============================] - 0s 9ms/step - loss: 0.5566 - accuracy: 0.7233\n",
      "[0.556574285030365, 0.7232620716094971]\n",
      "24/24 [==============================] - 0s 2ms/step\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.72      0.72      0.72       374\n",
      "           1       0.72      0.72      0.72       374\n",
      "\n",
      "    accuracy                           0.72       748\n",
      "   macro avg       0.72      0.72      0.72       748\n",
      "weighted avg       0.72      0.72      0.72       748\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:24:08.340498: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    }
   ],
   "source": [
    "y_preds = ANN(X_train, y_train, X_test, y_test, \"binary_crossentropy\", -1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1cb92bac-e04d-44ba-9e06-8a9501ad545a",
   "metadata": {},
   "source": [
    "# Method 2: Oversampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "cd2eeb24-7b7c-4d5d-b1ed-5a01b7a68161",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random over-sampling:\n",
      "0    5163\n",
      "1    5163\n",
      "Name: Churn, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# Oversample 1-class and concat the DataFrames of both classes\n",
    "df_class_1_over = df_class_1.sample(count_class_0, replace=True)\n",
    "df_test_over = pd.concat([df_class_0, df_class_1_over], axis=0)\n",
    "\n",
    "print(\"Random over-sampling:\")\n",
    "print(df_test_over.Churn.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "1550dcbc-3cb6-4564-9431-56364fe51c0b",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df_test_over.drop(\"Churn\", axis=\"columns\")\n",
    "y = df_test_over[\"Churn\"]\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=15, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "045d2b9b-6edd-479c-84b3-43be89261dff",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    4130\n",
       "0    4130\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of classes in training Data\n",
    "y_train.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "268d2d8b-18cf-4ad7-96b7-516953a585d5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "  9/259 [>.............................] - ETA: 1s - loss: 0.6987 - accuracy: 0.5000  "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:29:20.491646: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "259/259 [==============================] - 2s 7ms/step - loss: 0.5594 - accuracy: 0.7229\n",
      "Epoch 2/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4918 - accuracy: 0.7623\n",
      "Epoch 3/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4824 - accuracy: 0.7672\n",
      "Epoch 4/100\n",
      "259/259 [==============================] - 1s 5ms/step - loss: 0.4792 - accuracy: 0.7673\n",
      "Epoch 5/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4748 - accuracy: 0.7686\n",
      "Epoch 6/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4713 - accuracy: 0.7736\n",
      "Epoch 7/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4684 - accuracy: 0.7738\n",
      "Epoch 8/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4667 - accuracy: 0.7752\n",
      "Epoch 9/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4643 - accuracy: 0.7771\n",
      "Epoch 10/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4618 - accuracy: 0.7816\n",
      "Epoch 11/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4588 - accuracy: 0.7801\n",
      "Epoch 12/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4571 - accuracy: 0.7811\n",
      "Epoch 13/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4549 - accuracy: 0.7834\n",
      "Epoch 14/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4529 - accuracy: 0.7838\n",
      "Epoch 15/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4522 - accuracy: 0.7834\n",
      "Epoch 16/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4494 - accuracy: 0.7890\n",
      "Epoch 17/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4468 - accuracy: 0.7890\n",
      "Epoch 18/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4459 - accuracy: 0.7877\n",
      "Epoch 19/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4430 - accuracy: 0.7962\n",
      "Epoch 20/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4418 - accuracy: 0.7946\n",
      "Epoch 21/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4405 - accuracy: 0.7979\n",
      "Epoch 22/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4369 - accuracy: 0.7987\n",
      "Epoch 23/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4378 - accuracy: 0.7953\n",
      "Epoch 24/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4349 - accuracy: 0.7982\n",
      "Epoch 25/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4334 - accuracy: 0.7981\n",
      "Epoch 26/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4316 - accuracy: 0.8012\n",
      "Epoch 27/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4297 - accuracy: 0.8000\n",
      "Epoch 28/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4287 - accuracy: 0.8030\n",
      "Epoch 29/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4270 - accuracy: 0.7993\n",
      "Epoch 30/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4245 - accuracy: 0.8056\n",
      "Epoch 31/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4232 - accuracy: 0.8068\n",
      "Epoch 32/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4223 - accuracy: 0.8058\n",
      "Epoch 33/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4203 - accuracy: 0.8092\n",
      "Epoch 34/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4194 - accuracy: 0.8075\n",
      "Epoch 35/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4173 - accuracy: 0.8107\n",
      "Epoch 36/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4163 - accuracy: 0.8098\n",
      "Epoch 37/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4141 - accuracy: 0.8085\n",
      "Epoch 38/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4135 - accuracy: 0.8110\n",
      "Epoch 39/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4111 - accuracy: 0.8144\n",
      "Epoch 40/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4095 - accuracy: 0.8143\n",
      "Epoch 41/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4068 - accuracy: 0.8179\n",
      "Epoch 42/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4084 - accuracy: 0.8162\n",
      "Epoch 43/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4051 - accuracy: 0.8174\n",
      "Epoch 44/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.4048 - accuracy: 0.8165\n",
      "Epoch 45/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4033 - accuracy: 0.8188\n",
      "Epoch 46/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4027 - accuracy: 0.8197\n",
      "Epoch 47/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.4011 - accuracy: 0.8208\n",
      "Epoch 48/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3997 - accuracy: 0.8190\n",
      "Epoch 49/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3997 - accuracy: 0.8203\n",
      "Epoch 50/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3987 - accuracy: 0.8176\n",
      "Epoch 51/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3975 - accuracy: 0.8166\n",
      "Epoch 52/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3959 - accuracy: 0.8219\n",
      "Epoch 53/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3966 - accuracy: 0.8199\n",
      "Epoch 54/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3931 - accuracy: 0.8241\n",
      "Epoch 55/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3914 - accuracy: 0.8220\n",
      "Epoch 56/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3907 - accuracy: 0.8282\n",
      "Epoch 57/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3914 - accuracy: 0.8241\n",
      "Epoch 58/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3902 - accuracy: 0.8293\n",
      "Epoch 59/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3879 - accuracy: 0.8262\n",
      "Epoch 60/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3873 - accuracy: 0.8286\n",
      "Epoch 61/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3859 - accuracy: 0.8283\n",
      "Epoch 62/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3856 - accuracy: 0.8304\n",
      "Epoch 63/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3850 - accuracy: 0.8278\n",
      "Epoch 64/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3843 - accuracy: 0.8285\n",
      "Epoch 65/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3830 - accuracy: 0.8315\n",
      "Epoch 66/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3835 - accuracy: 0.8289\n",
      "Epoch 67/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3804 - accuracy: 0.8272\n",
      "Epoch 68/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3829 - accuracy: 0.8303\n",
      "Epoch 69/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3806 - accuracy: 0.8331\n",
      "Epoch 70/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3820 - accuracy: 0.8303\n",
      "Epoch 71/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3800 - accuracy: 0.8312\n",
      "Epoch 72/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3788 - accuracy: 0.8320\n",
      "Epoch 73/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3780 - accuracy: 0.8337\n",
      "Epoch 74/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3788 - accuracy: 0.8322\n",
      "Epoch 75/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3776 - accuracy: 0.8354\n",
      "Epoch 76/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3754 - accuracy: 0.8357\n",
      "Epoch 77/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3749 - accuracy: 0.8343\n",
      "Epoch 78/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3745 - accuracy: 0.8326\n",
      "Epoch 79/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3753 - accuracy: 0.8333\n",
      "Epoch 80/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3731 - accuracy: 0.8361\n",
      "Epoch 81/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3739 - accuracy: 0.8357\n",
      "Epoch 82/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3720 - accuracy: 0.8366\n",
      "Epoch 83/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3712 - accuracy: 0.8351\n",
      "Epoch 84/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3724 - accuracy: 0.8360\n",
      "Epoch 85/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3710 - accuracy: 0.8374\n",
      "Epoch 86/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3703 - accuracy: 0.8355\n",
      "Epoch 87/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3709 - accuracy: 0.8378\n",
      "Epoch 88/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3683 - accuracy: 0.8360\n",
      "Epoch 89/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3674 - accuracy: 0.8406\n",
      "Epoch 90/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3673 - accuracy: 0.8380\n",
      "Epoch 91/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3677 - accuracy: 0.8392\n",
      "Epoch 92/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3653 - accuracy: 0.8385\n",
      "Epoch 93/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3657 - accuracy: 0.8402\n",
      "Epoch 94/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3664 - accuracy: 0.8366\n",
      "Epoch 95/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3656 - accuracy: 0.8364\n",
      "Epoch 96/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3653 - accuracy: 0.8409\n",
      "Epoch 97/100\n",
      "259/259 [==============================] - 2s 6ms/step - loss: 0.3648 - accuracy: 0.8377\n",
      "Epoch 98/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3631 - accuracy: 0.8412\n",
      "Epoch 99/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3643 - accuracy: 0.8401\n",
      "Epoch 100/100\n",
      "259/259 [==============================] - 1s 6ms/step - loss: 0.3610 - accuracy: 0.8420\n",
      "28/65 [===========>..................] - ETA: 0s - loss: 0.4663 - accuracy: 0.7868"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:31:49.711621: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "65/65 [==============================] - 0s 4ms/step - loss: 0.4652 - accuracy: 0.7875\n",
      "[0.4651833474636078, 0.7875121235847473]\n",
      "65/65 [==============================] - 0s 2ms/step\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.79      0.78      0.79      1033\n",
      "           1       0.78      0.80      0.79      1033\n",
      "\n",
      "    accuracy                           0.79      2066\n",
      "   macro avg       0.79      0.79      0.79      2066\n",
      "weighted avg       0.79      0.79      0.79      2066\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:31:50.069526: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    }
   ],
   "source": [
    "loss = keras.losses.BinaryCrossentropy()\n",
    "weights = -1\n",
    "y_preds = ANN(X_train, y_train, X_test, y_test, \"binary_crossentropy\", -1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b32c1be-5715-4d53-bb18-ade31bd6cb70",
   "metadata": {},
   "source": [
    "# Method 3: SMOTE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "10be912b-1712-4ecd-9bd5-e73dfab4d120",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting imbalanced-learn\n",
      "  Downloading imbalanced_learn-0.10.1-py3-none-any.whl (226 kB)\n",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m226.0/226.0 kB\u001b[0m \u001b[31m3.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n",
      "\u001b[?25hRequirement already satisfied: threadpoolctl>=2.0.0 in /Users/serhatderya/miniconda3/envs/tfm1_arm64/lib/python3.9/site-packages (from imbalanced-learn) (2.2.0)\n",
      "Requirement already satisfied: scipy>=1.3.2 in /Users/serhatderya/miniconda3/envs/tfm1_arm64/lib/python3.9/site-packages (from imbalanced-learn) (1.9.3)\n",
      "Requirement already satisfied: joblib>=1.1.1 in /Users/serhatderya/miniconda3/envs/tfm1_arm64/lib/python3.9/site-packages (from imbalanced-learn) (1.1.1)\n",
      "Requirement already satisfied: scikit-learn>=1.0.2 in /Users/serhatderya/miniconda3/envs/tfm1_arm64/lib/python3.9/site-packages (from imbalanced-learn) (1.1.3)\n",
      "Requirement already satisfied: numpy>=1.17.3 in /Users/serhatderya/miniconda3/envs/tfm1_arm64/lib/python3.9/site-packages (from imbalanced-learn) (1.24.1)\n",
      "Installing collected packages: imbalanced-learn\n",
      "Successfully installed imbalanced-learn-0.10.1\n"
     ]
    }
   ],
   "source": [
    "!pip install imbalanced-learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "1b023e2c-44ee-4715-b1a8-0a846609d980",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df2.drop('Churn',axis='columns')\n",
    "y = df2['Churn']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2a5e05f4-aaae-45e3-92ae-e4e63355aef9",
   "metadata": {},
   "outputs": [],
   "source": [
    "from imblearn.over_sampling import SMOTE\n",
    "\n",
    "smote = SMOTE(sampling_strategy='minority')\n",
    "X_sm, y_sm = smote.fit_resample(X, y)\n",
    "\n",
    "y_sm.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5fbafe42-7b1f-48f2-bfb4-c50fbcf0c3a2",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_sm, y_sm, test_size=0.2, random_state=15, stratify=y_sm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a6dec5a8-cea5-4035-ab34-21988d584acb",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Number of classes in training Data\n",
    "y_train.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5891205e-823e-48a2-9ee0-f1401f281fcd",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_preds = ANN(X_train, y_train, X_test, y_test, 'binary_crossentropy', -1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e988da8-f827-4e54-b9a8-258a76783946",
   "metadata": {},
   "source": [
    "# Method 4: Use of Ensemble with undersampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "619a3637-4605-47e7-83c0-ec9d1a1a9add",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    5163\n",
       "1    1869\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.Churn.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "2f70ccc5-b3bc-41c1-987c-c9e7784db791",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Regain Original features and labels\n",
    "X = df2.drop(\"Churn\", axis=\"columns\")\n",
    "y = df2[\"Churn\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "cdd85473-123c-4b8e-88a7-421e232da0f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=15, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "3c16fba9-8876-463e-aff5-215fbd694ab4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    4130\n",
       "1    1495\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23356079-3c71-4914-ad9b-998ea456e791",
   "metadata": {},
   "source": [
    "#### model1 --> class1(1495) + class0(0,1495)\n",
    "#### model2 --> class1(1495) + class0(1496,2990)\n",
    "#### model3 --> class1(1495) + class0(2990,4130)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "de594bbc-7f4b-40dc-80f5-c1444ce474bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "df3 = X_train.copy()\n",
    "df3[\"Churn\"] = y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "662e9e08-0f97-45cd-8c78-6761994eaa5f",
   "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>gender</th>\n",
       "      <th>SeniorCitizen</th>\n",
       "      <th>Partner</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>tenure</th>\n",
       "      <th>PhoneService</th>\n",
       "      <th>MultipleLines</th>\n",
       "      <th>OnlineSecurity</th>\n",
       "      <th>OnlineBackup</th>\n",
       "      <th>DeviceProtection</th>\n",
       "      <th>...</th>\n",
       "      <th>InternetService_Fiber optic</th>\n",
       "      <th>InternetService_No</th>\n",
       "      <th>Contract_Month-to-month</th>\n",
       "      <th>Contract_One year</th>\n",
       "      <th>Contract_Two year</th>\n",
       "      <th>PaymentMethod_Bank transfer (automatic)</th>\n",
       "      <th>PaymentMethod_Credit card (automatic)</th>\n",
       "      <th>PaymentMethod_Electronic check</th>\n",
       "      <th>PaymentMethod_Mailed check</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>684</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</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>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2446</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.239437</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</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>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1680</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.774648</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2220</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2842</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.042254</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</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>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      gender  SeniorCitizen  Partner  Dependents    tenure  PhoneService  \\\n",
       "684        1              0        0           0  0.000000             1   \n",
       "2446       1              0        0           0  0.239437             1   \n",
       "1680       0              0        1           1  0.774648             1   \n",
       "2220       0              0        1           0  1.000000             1   \n",
       "2842       1              0        0           0  0.042254             0   \n",
       "\n",
       "      MultipleLines  OnlineSecurity  OnlineBackup  DeviceProtection  ...  \\\n",
       "684               0               0             0                 0  ...   \n",
       "2446              1               0             1                 0  ...   \n",
       "1680              1               0             0                 0  ...   \n",
       "2220              0               1             1                 0  ...   \n",
       "2842              0               1             0                 1  ...   \n",
       "\n",
       "      InternetService_Fiber optic  InternetService_No  \\\n",
       "684                             1                   0   \n",
       "2446                            1                   0   \n",
       "1680                            0                   1   \n",
       "2220                            0                   0   \n",
       "2842                            0                   0   \n",
       "\n",
       "      Contract_Month-to-month  Contract_One year  Contract_Two year  \\\n",
       "684                         1                  0                  0   \n",
       "2446                        1                  0                  0   \n",
       "1680                        0                  1                  0   \n",
       "2220                        0                  0                  1   \n",
       "2842                        1                  0                  0   \n",
       "\n",
       "      PaymentMethod_Bank transfer (automatic)  \\\n",
       "684                                         0   \n",
       "2446                                        0   \n",
       "1680                                        0   \n",
       "2220                                        1   \n",
       "2842                                        0   \n",
       "\n",
       "      PaymentMethod_Credit card (automatic)  PaymentMethod_Electronic check  \\\n",
       "684                                       0                               0   \n",
       "2446                                      1                               0   \n",
       "1680                                      0                               0   \n",
       "2220                                      0                               0   \n",
       "2842                                      0                               0   \n",
       "\n",
       "      PaymentMethod_Mailed check  Churn  \n",
       "684                            1      0  \n",
       "2446                           0      1  \n",
       "1680                           1      0  \n",
       "2220                           0      0  \n",
       "2842                           1      0  \n",
       "\n",
       "[5 rows x 27 columns]"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "8a47eafb-4f8c-4535-83cb-b93d4a553c29",
   "metadata": {},
   "outputs": [],
   "source": [
    "df3_class0 = df3[df3.Churn==0]\n",
    "df3_class1 = df3[df3.Churn==1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "f09eec43-9487-47e9-944a-2155c90419dc",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_train_batch(df_majority, df_minority, start, end):\n",
    "    df_train = pd.concat([df_majority[start:end], df_minority], axis=0)\n",
    "    \n",
    "    X_train = df_train.drop(\"Churn\", axis=\"columns\")\n",
    "    y_train = df_train.Churn\n",
    "    return X_train, y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "4966360d-cd05-4947-9712-30ddb048567b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      " 8/94 [=>............................] - ETA: 0s - loss: 0.7002 - accuracy: 0.4961 "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:46:11.539069: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "94/94 [==============================] - 1s 7ms/step - loss: 0.6380 - accuracy: 0.6525\n",
      "Epoch 2/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.5242 - accuracy: 0.7505\n",
      "Epoch 3/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4964 - accuracy: 0.7662\n",
      "Epoch 4/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4886 - accuracy: 0.7712\n",
      "Epoch 5/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4843 - accuracy: 0.7662\n",
      "Epoch 6/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4830 - accuracy: 0.7716\n",
      "Epoch 7/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4792 - accuracy: 0.7692\n",
      "Epoch 8/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4778 - accuracy: 0.7689\n",
      "Epoch 9/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4779 - accuracy: 0.7739\n",
      "Epoch 10/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4750 - accuracy: 0.7689\n",
      "Epoch 11/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4726 - accuracy: 0.7736\n",
      "Epoch 12/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4711 - accuracy: 0.7753\n",
      "Epoch 13/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4701 - accuracy: 0.7776\n",
      "Epoch 14/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4696 - accuracy: 0.7763\n",
      "Epoch 15/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4674 - accuracy: 0.7779\n",
      "Epoch 16/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4659 - accuracy: 0.7769\n",
      "Epoch 17/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4652 - accuracy: 0.7826\n",
      "Epoch 18/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4632 - accuracy: 0.7813\n",
      "Epoch 19/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4633 - accuracy: 0.7796\n",
      "Epoch 20/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4614 - accuracy: 0.7809\n",
      "Epoch 21/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4598 - accuracy: 0.7883\n",
      "Epoch 22/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4596 - accuracy: 0.7839\n",
      "Epoch 23/100\n",
      "94/94 [==============================] - 1s 7ms/step - loss: 0.4578 - accuracy: 0.7853\n",
      "Epoch 24/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4568 - accuracy: 0.7866\n",
      "Epoch 25/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4562 - accuracy: 0.7833\n",
      "Epoch 26/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4539 - accuracy: 0.7849\n",
      "Epoch 27/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4521 - accuracy: 0.7886\n",
      "Epoch 28/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4504 - accuracy: 0.7916\n",
      "Epoch 29/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4495 - accuracy: 0.7900\n",
      "Epoch 30/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4496 - accuracy: 0.7866\n",
      "Epoch 31/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4481 - accuracy: 0.7946\n",
      "Epoch 32/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4489 - accuracy: 0.7890\n",
      "Epoch 33/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4449 - accuracy: 0.7913\n",
      "Epoch 34/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4441 - accuracy: 0.7946\n",
      "Epoch 35/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4422 - accuracy: 0.7923\n",
      "Epoch 36/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4419 - accuracy: 0.7970\n",
      "Epoch 37/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4406 - accuracy: 0.7957\n",
      "Epoch 38/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4387 - accuracy: 0.7946\n",
      "Epoch 39/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4390 - accuracy: 0.8027\n",
      "Epoch 40/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4365 - accuracy: 0.7967\n",
      "Epoch 41/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4350 - accuracy: 0.7973\n",
      "Epoch 42/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4345 - accuracy: 0.8013\n",
      "Epoch 43/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4342 - accuracy: 0.7930\n",
      "Epoch 44/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4338 - accuracy: 0.7933\n",
      "Epoch 45/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4323 - accuracy: 0.7990\n",
      "Epoch 46/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4320 - accuracy: 0.7990\n",
      "Epoch 47/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4290 - accuracy: 0.8003\n",
      "Epoch 48/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4286 - accuracy: 0.8013\n",
      "Epoch 49/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4288 - accuracy: 0.8040\n",
      "Epoch 50/100\n",
      "94/94 [==============================] - 1s 7ms/step - loss: 0.4276 - accuracy: 0.8030\n",
      "Epoch 51/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4275 - accuracy: 0.7993\n",
      "Epoch 52/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4249 - accuracy: 0.8060\n",
      "Epoch 53/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4250 - accuracy: 0.8070\n",
      "Epoch 54/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4245 - accuracy: 0.8107\n",
      "Epoch 55/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4239 - accuracy: 0.8100\n",
      "Epoch 56/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4191 - accuracy: 0.8087\n",
      "Epoch 57/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4215 - accuracy: 0.8057\n",
      "Epoch 58/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4216 - accuracy: 0.8070\n",
      "Epoch 59/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4198 - accuracy: 0.8057\n",
      "Epoch 60/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4185 - accuracy: 0.8037\n",
      "Epoch 61/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4189 - accuracy: 0.8117\n",
      "Epoch 62/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4171 - accuracy: 0.8100\n",
      "Epoch 63/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4170 - accuracy: 0.8047\n",
      "Epoch 64/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4140 - accuracy: 0.8120\n",
      "Epoch 65/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4172 - accuracy: 0.8107\n",
      "Epoch 66/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4121 - accuracy: 0.8137\n",
      "Epoch 67/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4140 - accuracy: 0.8130\n",
      "Epoch 68/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4125 - accuracy: 0.8114\n",
      "Epoch 69/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4114 - accuracy: 0.8167\n",
      "Epoch 70/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4113 - accuracy: 0.8151\n",
      "Epoch 71/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4117 - accuracy: 0.8114\n",
      "Epoch 72/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4103 - accuracy: 0.8157\n",
      "Epoch 73/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4086 - accuracy: 0.8197\n",
      "Epoch 74/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4075 - accuracy: 0.8167\n",
      "Epoch 75/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4072 - accuracy: 0.8167\n",
      "Epoch 76/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4055 - accuracy: 0.8231\n",
      "Epoch 77/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4051 - accuracy: 0.8177\n",
      "Epoch 78/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4057 - accuracy: 0.8157\n",
      "Epoch 79/100\n",
      "94/94 [==============================] - 1s 7ms/step - loss: 0.4053 - accuracy: 0.8214\n",
      "Epoch 80/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4038 - accuracy: 0.8174\n",
      "Epoch 81/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4061 - accuracy: 0.8171\n",
      "Epoch 82/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4029 - accuracy: 0.8154\n",
      "Epoch 83/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4011 - accuracy: 0.8217\n",
      "Epoch 84/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4014 - accuracy: 0.8234\n",
      "Epoch 85/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4000 - accuracy: 0.8194\n",
      "Epoch 86/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4005 - accuracy: 0.8191\n",
      "Epoch 87/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3985 - accuracy: 0.8244\n",
      "Epoch 88/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3970 - accuracy: 0.8251\n",
      "Epoch 89/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3974 - accuracy: 0.8227\n",
      "Epoch 90/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3965 - accuracy: 0.8251\n",
      "Epoch 91/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3971 - accuracy: 0.8197\n",
      "Epoch 92/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3972 - accuracy: 0.8241\n",
      "Epoch 93/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3970 - accuracy: 0.8187\n",
      "Epoch 94/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3964 - accuracy: 0.8207\n",
      "Epoch 95/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3943 - accuracy: 0.8221\n",
      "Epoch 96/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3940 - accuracy: 0.8278\n",
      "Epoch 97/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3936 - accuracy: 0.8217\n",
      "Epoch 98/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3920 - accuracy: 0.8261\n",
      "Epoch 99/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3921 - accuracy: 0.8251\n",
      "Epoch 100/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3916 - accuracy: 0.8251\n",
      "25/44 [================>.............] - ETA: 0s - loss: 0.5586 - accuracy: 0.7412"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:47:07.990536: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44/44 [==============================] - 0s 5ms/step - loss: 0.5714 - accuracy: 0.7249\n",
      "[0.5713969469070435, 0.7249467372894287]\n",
      "44/44 [==============================] - 0s 2ms/step\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.89      0.71      0.79      1033\n",
      "           1       0.49      0.76      0.59       374\n",
      "\n",
      "    accuracy                           0.72      1407\n",
      "   macro avg       0.69      0.74      0.69      1407\n",
      "weighted avg       0.78      0.72      0.74      1407\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:47:08.299038: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    }
   ],
   "source": [
    "X_train, y_train = get_train_batch(df3_class0, df3_class1, 0, 1495)\n",
    "\n",
    "y_pred1 = ANN(X_train, y_train, X_test, y_test, \"binary_crossentropy\", -1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "ec840c33-26c4-4ab1-8899-65d1c7c48695",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      " 7/94 [=>............................] - ETA: 0s - loss: 0.7074 - accuracy: 0.4330 "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:47:08.580961: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "94/94 [==============================] - 1s 7ms/step - loss: 0.6384 - accuracy: 0.6572\n",
      "Epoch 2/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.5231 - accuracy: 0.7565\n",
      "Epoch 3/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4987 - accuracy: 0.7622\n",
      "Epoch 4/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4900 - accuracy: 0.7625\n",
      "Epoch 5/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4842 - accuracy: 0.7659\n",
      "Epoch 6/100\n",
      "94/94 [==============================] - 1s 7ms/step - loss: 0.4813 - accuracy: 0.7696\n",
      "Epoch 7/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4765 - accuracy: 0.7696\n",
      "Epoch 8/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4744 - accuracy: 0.7676\n",
      "Epoch 9/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4712 - accuracy: 0.7706\n",
      "Epoch 10/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4717 - accuracy: 0.7692\n",
      "Epoch 11/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4679 - accuracy: 0.7742\n",
      "Epoch 12/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4644 - accuracy: 0.7686\n",
      "Epoch 13/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4630 - accuracy: 0.7739\n",
      "Epoch 14/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4606 - accuracy: 0.7786\n",
      "Epoch 15/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4600 - accuracy: 0.7749\n",
      "Epoch 16/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4578 - accuracy: 0.7786\n",
      "Epoch 17/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4567 - accuracy: 0.7846\n",
      "Epoch 18/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4548 - accuracy: 0.7786\n",
      "Epoch 19/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4538 - accuracy: 0.7786\n",
      "Epoch 20/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4515 - accuracy: 0.7846\n",
      "Epoch 21/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4511 - accuracy: 0.7843\n",
      "Epoch 22/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4491 - accuracy: 0.7833\n",
      "Epoch 23/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4481 - accuracy: 0.7823\n",
      "Epoch 24/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4465 - accuracy: 0.7883\n",
      "Epoch 25/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4438 - accuracy: 0.7910\n",
      "Epoch 26/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4430 - accuracy: 0.7880\n",
      "Epoch 27/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4434 - accuracy: 0.7886\n",
      "Epoch 28/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4398 - accuracy: 0.7886\n",
      "Epoch 29/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4400 - accuracy: 0.7920\n",
      "Epoch 30/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4375 - accuracy: 0.7960\n",
      "Epoch 31/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4364 - accuracy: 0.7967\n",
      "Epoch 32/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4353 - accuracy: 0.7933\n",
      "Epoch 33/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4340 - accuracy: 0.7950\n",
      "Epoch 34/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4347 - accuracy: 0.7946\n",
      "Epoch 35/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4331 - accuracy: 0.7963\n",
      "Epoch 36/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4314 - accuracy: 0.7953\n",
      "Epoch 37/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4316 - accuracy: 0.7933\n",
      "Epoch 38/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4302 - accuracy: 0.7967\n",
      "Epoch 39/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4271 - accuracy: 0.7990\n",
      "Epoch 40/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4271 - accuracy: 0.7960\n",
      "Epoch 41/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4277 - accuracy: 0.7993\n",
      "Epoch 42/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4260 - accuracy: 0.7977\n",
      "Epoch 43/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4243 - accuracy: 0.8000\n",
      "Epoch 44/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4220 - accuracy: 0.7990\n",
      "Epoch 45/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4217 - accuracy: 0.8010\n",
      "Epoch 46/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4202 - accuracy: 0.8030\n",
      "Epoch 47/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4211 - accuracy: 0.8010\n",
      "Epoch 48/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4195 - accuracy: 0.8023\n",
      "Epoch 49/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4170 - accuracy: 0.8040\n",
      "Epoch 50/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4165 - accuracy: 0.8077\n",
      "Epoch 51/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4142 - accuracy: 0.8054\n",
      "Epoch 52/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4156 - accuracy: 0.8060\n",
      "Epoch 53/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4138 - accuracy: 0.8043\n",
      "Epoch 54/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4124 - accuracy: 0.8104\n",
      "Epoch 55/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4114 - accuracy: 0.8050\n",
      "Epoch 56/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4122 - accuracy: 0.8087\n",
      "Epoch 57/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4085 - accuracy: 0.8077\n",
      "Epoch 58/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4098 - accuracy: 0.8077\n",
      "Epoch 59/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4074 - accuracy: 0.8100\n",
      "Epoch 60/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4053 - accuracy: 0.8124\n",
      "Epoch 61/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4052 - accuracy: 0.8110\n",
      "Epoch 62/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4049 - accuracy: 0.8094\n",
      "Epoch 63/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4025 - accuracy: 0.8090\n",
      "Epoch 64/100\n",
      "94/94 [==============================] - 1s 7ms/step - loss: 0.4038 - accuracy: 0.8140\n",
      "Epoch 65/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4008 - accuracy: 0.8167\n",
      "Epoch 66/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4003 - accuracy: 0.8094\n",
      "Epoch 67/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.4010 - accuracy: 0.8181\n",
      "Epoch 68/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3983 - accuracy: 0.8124\n",
      "Epoch 69/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3972 - accuracy: 0.8197\n",
      "Epoch 70/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3955 - accuracy: 0.8167\n",
      "Epoch 71/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3953 - accuracy: 0.8204\n",
      "Epoch 72/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3984 - accuracy: 0.8174\n",
      "Epoch 73/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3960 - accuracy: 0.8194\n",
      "Epoch 74/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3933 - accuracy: 0.8164\n",
      "Epoch 75/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3922 - accuracy: 0.8181\n",
      "Epoch 76/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3915 - accuracy: 0.8217\n",
      "Epoch 77/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3907 - accuracy: 0.8197\n",
      "Epoch 78/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3899 - accuracy: 0.8174\n",
      "Epoch 79/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3904 - accuracy: 0.8177\n",
      "Epoch 80/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3873 - accuracy: 0.8171\n",
      "Epoch 81/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3867 - accuracy: 0.8187\n",
      "Epoch 82/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3855 - accuracy: 0.8201\n",
      "Epoch 83/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3864 - accuracy: 0.8241\n",
      "Epoch 84/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3858 - accuracy: 0.8241\n",
      "Epoch 85/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3835 - accuracy: 0.8194\n",
      "Epoch 86/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3844 - accuracy: 0.8234\n",
      "Epoch 87/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3828 - accuracy: 0.8247\n",
      "Epoch 88/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3832 - accuracy: 0.8211\n",
      "Epoch 89/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3844 - accuracy: 0.8217\n",
      "Epoch 90/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3798 - accuracy: 0.8281\n",
      "Epoch 91/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3802 - accuracy: 0.8278\n",
      "Epoch 92/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3780 - accuracy: 0.8241\n",
      "Epoch 93/100\n",
      "94/94 [==============================] - 1s 7ms/step - loss: 0.3768 - accuracy: 0.8301\n",
      "Epoch 94/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3780 - accuracy: 0.8231\n",
      "Epoch 95/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3735 - accuracy: 0.8284\n",
      "Epoch 96/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3767 - accuracy: 0.8247\n",
      "Epoch 97/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3740 - accuracy: 0.8311\n",
      "Epoch 98/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3734 - accuracy: 0.8237\n",
      "Epoch 99/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3742 - accuracy: 0.8288\n",
      "Epoch 100/100\n",
      "94/94 [==============================] - 1s 6ms/step - loss: 0.3735 - accuracy: 0.8281\n",
      "27/44 [=================>............] - ETA: 0s - loss: 0.6128 - accuracy: 0.7049"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:48:03.917530: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44/44 [==============================] - 0s 5ms/step - loss: 0.6198 - accuracy: 0.7050\n",
      "[0.6197667717933655, 0.7050462365150452]\n",
      "44/44 [==============================] - 0s 2ms/step\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.89      0.68      0.77      1033\n",
      "           1       0.47      0.76      0.58       374\n",
      "\n",
      "    accuracy                           0.71      1407\n",
      "   macro avg       0.68      0.72      0.68      1407\n",
      "weighted avg       0.78      0.71      0.72      1407\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:48:04.211870: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    }
   ],
   "source": [
    "X_train, y_train = get_train_batch(df3_class0, df3_class1, 1495, 2990)\n",
    "\n",
    "y_pred2 = ANN(X_train, y_train, X_test, y_test, 'binary_crossentropy', -1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "99fee6ba-dab8-4fff-a863-5e6df9cdd637",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      " 7/83 [=>............................] - ETA: 0s - loss: 0.6862 - accuracy: 0.5357 "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:48:04.490548: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "83/83 [==============================] - 1s 10ms/step - loss: 0.6401 - accuracy: 0.6315\n",
      "Epoch 2/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.5339 - accuracy: 0.7560\n",
      "Epoch 3/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4976 - accuracy: 0.7681\n",
      "Epoch 4/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4863 - accuracy: 0.7662\n",
      "Epoch 5/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4801 - accuracy: 0.7693\n",
      "Epoch 6/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4767 - accuracy: 0.7753\n",
      "Epoch 7/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4735 - accuracy: 0.7772\n",
      "Epoch 8/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4688 - accuracy: 0.7765\n",
      "Epoch 9/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4672 - accuracy: 0.7822\n",
      "Epoch 10/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4643 - accuracy: 0.7829\n",
      "Epoch 11/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4625 - accuracy: 0.7856\n",
      "Epoch 12/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4603 - accuracy: 0.7837\n",
      "Epoch 13/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4587 - accuracy: 0.7852\n",
      "Epoch 14/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4556 - accuracy: 0.7879\n",
      "Epoch 15/100\n",
      "83/83 [==============================] - 1s 6ms/step - loss: 0.4551 - accuracy: 0.7875\n",
      "Epoch 16/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4524 - accuracy: 0.7898\n",
      "Epoch 17/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4524 - accuracy: 0.7886\n",
      "Epoch 18/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4490 - accuracy: 0.7905\n",
      "Epoch 19/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4469 - accuracy: 0.7947\n",
      "Epoch 20/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4466 - accuracy: 0.7958\n",
      "Epoch 21/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4438 - accuracy: 0.7901\n",
      "Epoch 22/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4419 - accuracy: 0.7932\n",
      "Epoch 23/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4390 - accuracy: 0.7951\n",
      "Epoch 24/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4396 - accuracy: 0.7962\n",
      "Epoch 25/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4380 - accuracy: 0.8000\n",
      "Epoch 26/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4364 - accuracy: 0.8004\n",
      "Epoch 27/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4354 - accuracy: 0.7996\n",
      "Epoch 28/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4318 - accuracy: 0.8015\n",
      "Epoch 29/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4307 - accuracy: 0.8008\n",
      "Epoch 30/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4298 - accuracy: 0.8008\n",
      "Epoch 31/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4277 - accuracy: 0.8011\n",
      "Epoch 32/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4283 - accuracy: 0.8004\n",
      "Epoch 33/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4255 - accuracy: 0.8057\n",
      "Epoch 34/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4261 - accuracy: 0.8042\n",
      "Epoch 35/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4225 - accuracy: 0.8102\n",
      "Epoch 36/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4219 - accuracy: 0.8065\n",
      "Epoch 37/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4214 - accuracy: 0.8053\n",
      "Epoch 38/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4186 - accuracy: 0.8083\n",
      "Epoch 39/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4185 - accuracy: 0.8102\n",
      "Epoch 40/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4170 - accuracy: 0.8099\n",
      "Epoch 41/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4163 - accuracy: 0.8114\n",
      "Epoch 42/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4167 - accuracy: 0.8144\n",
      "Epoch 43/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4121 - accuracy: 0.8114\n",
      "Epoch 44/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4120 - accuracy: 0.8129\n",
      "Epoch 45/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4124 - accuracy: 0.8118\n",
      "Epoch 46/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4138 - accuracy: 0.8121\n",
      "Epoch 47/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4110 - accuracy: 0.8137\n",
      "Epoch 48/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4084 - accuracy: 0.8167\n",
      "Epoch 49/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4056 - accuracy: 0.8152\n",
      "Epoch 50/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4081 - accuracy: 0.8167\n",
      "Epoch 51/100\n",
      "83/83 [==============================] - 1s 6ms/step - loss: 0.4058 - accuracy: 0.8140\n",
      "Epoch 52/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4031 - accuracy: 0.8190\n",
      "Epoch 53/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4045 - accuracy: 0.8156\n",
      "Epoch 54/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4035 - accuracy: 0.8197\n",
      "Epoch 55/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.4023 - accuracy: 0.8182\n",
      "Epoch 56/100\n",
      "83/83 [==============================] - 1s 6ms/step - loss: 0.3990 - accuracy: 0.8231\n",
      "Epoch 57/100\n",
      "83/83 [==============================] - 1s 6ms/step - loss: 0.3992 - accuracy: 0.8216\n",
      "Epoch 58/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3981 - accuracy: 0.8194\n",
      "Epoch 59/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3992 - accuracy: 0.8216\n",
      "Epoch 60/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3982 - accuracy: 0.8159\n",
      "Epoch 61/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3960 - accuracy: 0.8224\n",
      "Epoch 62/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3954 - accuracy: 0.8205\n",
      "Epoch 63/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3947 - accuracy: 0.8250\n",
      "Epoch 64/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3926 - accuracy: 0.8262\n",
      "Epoch 65/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3923 - accuracy: 0.8247\n",
      "Epoch 66/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3912 - accuracy: 0.8224\n",
      "Epoch 67/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3922 - accuracy: 0.8273\n",
      "Epoch 68/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3914 - accuracy: 0.8258\n",
      "Epoch 69/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3907 - accuracy: 0.8239\n",
      "Epoch 70/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3893 - accuracy: 0.8194\n",
      "Epoch 71/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3881 - accuracy: 0.8304\n",
      "Epoch 72/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3861 - accuracy: 0.8288\n",
      "Epoch 73/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3874 - accuracy: 0.8266\n",
      "Epoch 74/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3838 - accuracy: 0.8319\n",
      "Epoch 75/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3844 - accuracy: 0.8269\n",
      "Epoch 76/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3841 - accuracy: 0.8266\n",
      "Epoch 77/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3832 - accuracy: 0.8296\n",
      "Epoch 78/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3820 - accuracy: 0.8311\n",
      "Epoch 79/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3815 - accuracy: 0.8277\n",
      "Epoch 80/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3805 - accuracy: 0.8304\n",
      "Epoch 81/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3795 - accuracy: 0.8292\n",
      "Epoch 82/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3804 - accuracy: 0.8285\n",
      "Epoch 83/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3771 - accuracy: 0.8285\n",
      "Epoch 84/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3787 - accuracy: 0.8307\n",
      "Epoch 85/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3798 - accuracy: 0.8262\n",
      "Epoch 86/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3755 - accuracy: 0.8311\n",
      "Epoch 87/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3751 - accuracy: 0.8296\n",
      "Epoch 88/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3755 - accuracy: 0.8315\n",
      "Epoch 89/100\n",
      "83/83 [==============================] - 1s 6ms/step - loss: 0.3742 - accuracy: 0.8326\n",
      "Epoch 90/100\n",
      "83/83 [==============================] - 1s 6ms/step - loss: 0.3744 - accuracy: 0.8345\n",
      "Epoch 91/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3715 - accuracy: 0.8349\n",
      "Epoch 92/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3710 - accuracy: 0.8364\n",
      "Epoch 93/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3697 - accuracy: 0.8376\n",
      "Epoch 94/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3714 - accuracy: 0.8326\n",
      "Epoch 95/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3712 - accuracy: 0.8319\n",
      "Epoch 96/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3688 - accuracy: 0.8414\n",
      "Epoch 97/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3675 - accuracy: 0.8372\n",
      "Epoch 98/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3668 - accuracy: 0.8364\n",
      "Epoch 99/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3655 - accuracy: 0.8398\n",
      "Epoch 100/100\n",
      "83/83 [==============================] - 0s 6ms/step - loss: 0.3661 - accuracy: 0.8376\n",
      "25/44 [================>.............] - ETA: 0s - loss: 0.6133 - accuracy: 0.6812"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:48:53.331973: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44/44 [==============================] - 0s 5ms/step - loss: 0.6346 - accuracy: 0.6823\n",
      "[0.6346449851989746, 0.6823027729988098]\n",
      "44/44 [==============================] - 0s 2ms/step\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.89      0.64      0.75      1033\n",
      "           1       0.44      0.79      0.57       374\n",
      "\n",
      "    accuracy                           0.68      1407\n",
      "   macro avg       0.67      0.72      0.66      1407\n",
      "weighted avg       0.77      0.68      0.70      1407\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-01-20 09:48:53.640859: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:113] Plugin optimizer for device_type GPU is enabled.\n"
     ]
    }
   ],
   "source": [
    "X_train, y_train = get_train_batch(df3_class0, df3_class1, 2990, 4130)\n",
    "\n",
    "y_pred3 = ANN(X_train, y_train, X_test, y_test, 'binary_crossentropy', -1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "b5f46194-43bf-4088-b2ef-b5d1adffb003",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1407"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(y_pred1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "659c6575-2b34-4dd5-90cf-935a6cd119c1",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_final = y_pred1.copy()\n",
    "for i in range(len(y_pred1)):\n",
    "    n_ones = y_pred1[i] + y_pred2[i] + y_pred3[i]\n",
    "    if n_ones>1:\n",
    "        y_pred_final[i] = 1\n",
    "    else:\n",
    "        y_pred_final[i] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "aaffb930-58b3-42c8-89d2-8aa1308c1091",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.89      0.68      0.77      1033\n",
      "           1       0.47      0.78      0.59       374\n",
      "\n",
      "    accuracy                           0.71      1407\n",
      "   macro avg       0.68      0.73      0.68      1407\n",
      "weighted avg       0.78      0.71      0.72      1407\n",
      "\n"
     ]
    }
   ],
   "source": [
    "cl_rep = classification_report(y_test, y_pred_final)\n",
    "print(cl_rep)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a75ca688-0128-4fa0-b42a-3b95381418bf",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}