Bank Demo - Machine Learning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bank Demo"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Datasource from Kaggle : https://www.kaggle.com/lovelesh/bank-dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Business Scenario\n",
    "Given historic information of Customers Exiting the bank and their information -\n",
    "1. We want to identify the factors that have impact on Customers Exiting the Bank\n",
    "\n",
    "2. Build a Prediction Model to identify Customer who have higher probability of Exiting and use this data to plan for future Actions.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Analysis and Preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing Libraries\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 14)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Importing Dataset\n",
    "bank_data = pd.read_csv('C:/AI_DAS/Bank_Demo/Bank_Data.csv')\n",
    "bank_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>RowNumber</th>\n",
       "      <th>CustomerId</th>\n",
       "      <th>Surname</th>\n",
       "      <th>CreditScore</th>\n",
       "      <th>Geography</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure</th>\n",
       "      <th>Balance</th>\n",
       "      <th>NumOfProducts</th>\n",
       "      <th>HasCrCard</th>\n",
       "      <th>IsActiveMember</th>\n",
       "      <th>EstimatedSalary</th>\n",
       "      <th>Exited</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>15634602</td>\n",
       "      <td>Hargrave</td>\n",
       "      <td>619</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>42</td>\n",
       "      <td>2</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>101348.88</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>15647311</td>\n",
       "      <td>Hill</td>\n",
       "      <td>608</td>\n",
       "      <td>Spain</td>\n",
       "      <td>Female</td>\n",
       "      <td>41</td>\n",
       "      <td>1</td>\n",
       "      <td>83807.86</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>112542.58</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>15619304</td>\n",
       "      <td>Onio</td>\n",
       "      <td>502</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>42</td>\n",
       "      <td>8</td>\n",
       "      <td>159660.80</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113931.57</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>15701354</td>\n",
       "      <td>Boni</td>\n",
       "      <td>699</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>39</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>93826.63</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>15737888</td>\n",
       "      <td>Mitchell</td>\n",
       "      <td>850</td>\n",
       "      <td>Spain</td>\n",
       "      <td>Female</td>\n",
       "      <td>43</td>\n",
       "      <td>2</td>\n",
       "      <td>125510.82</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>79084.10</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   RowNumber  CustomerId   Surname  CreditScore Geography  Gender  Age  \\\n",
       "0          1    15634602  Hargrave          619    France  Female   42   \n",
       "1          2    15647311      Hill          608     Spain  Female   41   \n",
       "2          3    15619304      Onio          502    France  Female   42   \n",
       "3          4    15701354      Boni          699    France  Female   39   \n",
       "4          5    15737888  Mitchell          850     Spain  Female   43   \n",
       "\n",
       "   Tenure    Balance  NumOfProducts  HasCrCard  IsActiveMember  \\\n",
       "0       2       0.00              1          1               1   \n",
       "1       1   83807.86              1          0               1   \n",
       "2       8  159660.80              3          1               0   \n",
       "3       1       0.00              2          0               0   \n",
       "4       2  125510.82              1          1               1   \n",
       "\n",
       "   EstimatedSalary  Exited  \n",
       "0        101348.88       1  \n",
       "1        112542.58       0  \n",
       "2        113931.57       1  \n",
       "3         93826.63       0  \n",
       "4         79084.10       0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bank_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 10000 entries, 0 to 9999\n",
      "Data columns (total 14 columns):\n",
      "RowNumber          10000 non-null int64\n",
      "CustomerId         10000 non-null int64\n",
      "Surname            10000 non-null object\n",
      "CreditScore        10000 non-null int64\n",
      "Geography          10000 non-null object\n",
      "Gender             10000 non-null object\n",
      "Age                10000 non-null int64\n",
      "Tenure             10000 non-null int64\n",
      "Balance            10000 non-null float64\n",
      "NumOfProducts      10000 non-null int64\n",
      "HasCrCard          10000 non-null int64\n",
      "IsActiveMember     10000 non-null int64\n",
      "EstimatedSalary    10000 non-null float64\n",
      "Exited             10000 non-null int64\n",
      "dtypes: float64(2), int64(9), object(3)\n",
      "memory usage: 1.1+ MB\n"
     ]
    }
   ],
   "source": [
    "# Checking data schema\n",
    "data = bank_data.copy()\n",
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RowNumber          0\n",
       "CustomerId         0\n",
       "Surname            0\n",
       "CreditScore        0\n",
       "Geography          0\n",
       "Gender             0\n",
       "Age                0\n",
       "Tenure             0\n",
       "Balance            0\n",
       "NumOfProducts      0\n",
       "HasCrCard          0\n",
       "IsActiveMember     0\n",
       "EstimatedSalary    0\n",
       "Exited             0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check for Null Values\n",
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since all columns are non-null, there is no missing data values. If data has missing information, then we need to replace non-null values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RowNumber          10000\n",
       "CustomerId         10000\n",
       "Surname             2932\n",
       "CreditScore          460\n",
       "Geography              3\n",
       "Gender                 2\n",
       "Age                   70\n",
       "Tenure                11\n",
       "Balance             6382\n",
       "NumOfProducts          4\n",
       "HasCrCard              2\n",
       "IsActiveMember         2\n",
       "EstimatedSalary     9999\n",
       "Exited                 2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get unique count for each variable\n",
    "data.nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>RowNumber</th>\n",
       "      <th>CustomerId</th>\n",
       "      <th>CreditScore</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure</th>\n",
       "      <th>Balance</th>\n",
       "      <th>NumOfProducts</th>\n",
       "      <th>HasCrCard</th>\n",
       "      <th>IsActiveMember</th>\n",
       "      <th>EstimatedSalary</th>\n",
       "      <th>Exited</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000.00000</td>\n",
       "      <td>1.000000e+04</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.00000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>5000.50000</td>\n",
       "      <td>1.569094e+07</td>\n",
       "      <td>650.528800</td>\n",
       "      <td>38.921800</td>\n",
       "      <td>5.012800</td>\n",
       "      <td>76485.889288</td>\n",
       "      <td>1.530200</td>\n",
       "      <td>0.70550</td>\n",
       "      <td>0.515100</td>\n",
       "      <td>100090.239881</td>\n",
       "      <td>0.203700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2886.89568</td>\n",
       "      <td>7.193619e+04</td>\n",
       "      <td>96.653299</td>\n",
       "      <td>10.487806</td>\n",
       "      <td>2.892174</td>\n",
       "      <td>62397.405202</td>\n",
       "      <td>0.581654</td>\n",
       "      <td>0.45584</td>\n",
       "      <td>0.499797</td>\n",
       "      <td>57510.492818</td>\n",
       "      <td>0.402769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.00000</td>\n",
       "      <td>1.556570e+07</td>\n",
       "      <td>350.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>11.580000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2500.75000</td>\n",
       "      <td>1.562853e+07</td>\n",
       "      <td>584.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>51002.110000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>5000.50000</td>\n",
       "      <td>1.569074e+07</td>\n",
       "      <td>652.000000</td>\n",
       "      <td>37.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>97198.540000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>100193.915000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7500.25000</td>\n",
       "      <td>1.575323e+07</td>\n",
       "      <td>718.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>127644.240000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>149388.247500</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>10000.00000</td>\n",
       "      <td>1.581569e+07</td>\n",
       "      <td>850.000000</td>\n",
       "      <td>92.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>250898.090000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>199992.480000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         RowNumber    CustomerId   CreditScore           Age        Tenure  \\\n",
       "count  10000.00000  1.000000e+04  10000.000000  10000.000000  10000.000000   \n",
       "mean    5000.50000  1.569094e+07    650.528800     38.921800      5.012800   \n",
       "std     2886.89568  7.193619e+04     96.653299     10.487806      2.892174   \n",
       "min        1.00000  1.556570e+07    350.000000     18.000000      0.000000   \n",
       "25%     2500.75000  1.562853e+07    584.000000     32.000000      3.000000   \n",
       "50%     5000.50000  1.569074e+07    652.000000     37.000000      5.000000   \n",
       "75%     7500.25000  1.575323e+07    718.000000     44.000000      7.000000   \n",
       "max    10000.00000  1.581569e+07    850.000000     92.000000     10.000000   \n",
       "\n",
       "             Balance  NumOfProducts    HasCrCard  IsActiveMember  \\\n",
       "count   10000.000000   10000.000000  10000.00000    10000.000000   \n",
       "mean    76485.889288       1.530200      0.70550        0.515100   \n",
       "std     62397.405202       0.581654      0.45584        0.499797   \n",
       "min         0.000000       1.000000      0.00000        0.000000   \n",
       "25%         0.000000       1.000000      0.00000        0.000000   \n",
       "50%     97198.540000       1.000000      1.00000        1.000000   \n",
       "75%    127644.240000       2.000000      1.00000        1.000000   \n",
       "max    250898.090000       4.000000      1.00000        1.000000   \n",
       "\n",
       "       EstimatedSalary        Exited  \n",
       "count     10000.000000  10000.000000  \n",
       "mean     100090.239881      0.203700  \n",
       "std       57510.492818      0.402769  \n",
       "min          11.580000      0.000000  \n",
       "25%       51002.110000      0.000000  \n",
       "50%      100193.915000      0.000000  \n",
       "75%      149388.247500      0.000000  \n",
       "max      199992.480000      1.000000  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# checking Statistical information of dataset\n",
    "data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Exploration"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### We are analysing on Factors that affect Exiting of Customer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Percentage of Bank Customers who Exited Vs Continued"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "status = 'Exited', 'Continued'\n",
    "count_exit = data.Exited[data['Exited']==1].count()\n",
    "count_cont = data.Exited[data['Exited']==0].count()\n",
    "sizes = (count_exit, count_cont)\n",
    "explode = (0, 0.1)\n",
    "colors = ['lightcoral', 'skyblue', 'yellowgreen']\n",
    "fig1, ax1 = plt.subplots(figsize=(9,6))\n",
    "ax1.pie(sizes, explode=explode, labels=status, autopct='%1.1f%%',colors=colors, shadow=True, startangle=140)\n",
    "ax1.axis('equal')\n",
    "plt.title(\"Percentage of customer Exited Vs Continued\", size = 10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Ratio of Customers Exited Vs Continued by Factors - Geography, Gender, HasCrCard, IsActiveMember"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x25a0e69def0>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x864 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axarr = plt.subplots(2, 2, figsize=(20, 12))\n",
    "sns.countplot(x='Geography', hue = 'Exited',data = data, ax=axarr[0][0], palette = \"seismic\")\n",
    "sns.countplot(x='Gender', hue = 'Exited',data = data, ax=axarr[0][1], palette = \"seismic\")\n",
    "sns.countplot(x='HasCrCard', hue = 'Exited',data = data, ax=axarr[1][0], palette = \"seismic\")\n",
    "sns.countplot(x='IsActiveMember', hue = 'Exited',data = data, ax=axarr[1][1], palette = \"seismic\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Trend of Customers exiting by Age Factor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x25a0ef4a940>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data[\"Age\"].value_counts().plot.bar(figsize=(20,6), x= \"Age\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Preprocessing (Feature Engineering)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### Machine Learning Algorithm cannot use text data in data processing so non-numeric column values are either dropped if not significant or converted to Integer value.\n",
    "###### Categorical features - Surname, Geography, Gender\n",
    "###### Columns Not Relevant for Predcition - RowNumber, Surname, CustomerID\n",
    "###### Columns converted to Numeric Features - Geography, Gender"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Deleting Surname feature as its irrelevant for analysis\n",
    "data = data.drop(\"Surname\", axis=1)\n",
    "data = data.drop(\"RowNumber\", axis = 1)\n",
    "data = data.drop(\"CustomerId\", axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>CreditScore</th>\n",
       "      <th>Geography</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure</th>\n",
       "      <th>Balance</th>\n",
       "      <th>NumOfProducts</th>\n",
       "      <th>HasCrCard</th>\n",
       "      <th>IsActiveMember</th>\n",
       "      <th>EstimatedSalary</th>\n",
       "      <th>Exited</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>619</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>42</td>\n",
       "      <td>2</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>101348.88</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>608</td>\n",
       "      <td>Spain</td>\n",
       "      <td>Female</td>\n",
       "      <td>41</td>\n",
       "      <td>1</td>\n",
       "      <td>83807.86</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>112542.58</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>502</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>42</td>\n",
       "      <td>8</td>\n",
       "      <td>159660.80</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113931.57</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>699</td>\n",
       "      <td>France</td>\n",
       "      <td>Female</td>\n",
       "      <td>39</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>93826.63</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>850</td>\n",
       "      <td>Spain</td>\n",
       "      <td>Female</td>\n",
       "      <td>43</td>\n",
       "      <td>2</td>\n",
       "      <td>125510.82</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>79084.10</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   CreditScore Geography  Gender  Age  Tenure    Balance  NumOfProducts  \\\n",
       "0          619    France  Female   42       2       0.00              1   \n",
       "1          608     Spain  Female   41       1   83807.86              1   \n",
       "2          502    France  Female   42       8  159660.80              3   \n",
       "3          699    France  Female   39       1       0.00              2   \n",
       "4          850     Spain  Female   43       2  125510.82              1   \n",
       "\n",
       "   HasCrCard  IsActiveMember  EstimatedSalary  Exited  \n",
       "0          1               1        101348.88       1  \n",
       "1          0               1        112542.58       0  \n",
       "2          1               0        113931.57       1  \n",
       "3          0               0         93826.63       0  \n",
       "4          1               1         79084.10       0  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['France' 'Spain' 'Germany']\n",
      "['Female' 'Male']\n"
     ]
    }
   ],
   "source": [
    "print(data[\"Geography\"].unique())\n",
    "print(data[\"Gender\"].unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Mapping Gender to Integer\n",
    "gender = {'Male': 1,'Female': 2} \n",
    "data.Gender = [gender[item] for item in data.Gender] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>CreditScore</th>\n",
       "      <th>Geography</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure</th>\n",
       "      <th>Balance</th>\n",
       "      <th>NumOfProducts</th>\n",
       "      <th>HasCrCard</th>\n",
       "      <th>IsActiveMember</th>\n",
       "      <th>EstimatedSalary</th>\n",
       "      <th>Exited</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>619</td>\n",
       "      <td>France</td>\n",
       "      <td>2</td>\n",
       "      <td>42</td>\n",
       "      <td>2</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>101348.88</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>608</td>\n",
       "      <td>Spain</td>\n",
       "      <td>2</td>\n",
       "      <td>41</td>\n",
       "      <td>1</td>\n",
       "      <td>83807.86</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>112542.58</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>502</td>\n",
       "      <td>France</td>\n",
       "      <td>2</td>\n",
       "      <td>42</td>\n",
       "      <td>8</td>\n",
       "      <td>159660.80</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113931.57</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>699</td>\n",
       "      <td>France</td>\n",
       "      <td>2</td>\n",
       "      <td>39</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>93826.63</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>850</td>\n",
       "      <td>Spain</td>\n",
       "      <td>2</td>\n",
       "      <td>43</td>\n",
       "      <td>2</td>\n",
       "      <td>125510.82</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>79084.10</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   CreditScore Geography  Gender  Age  Tenure    Balance  NumOfProducts  \\\n",
       "0          619    France       2   42       2       0.00              1   \n",
       "1          608     Spain       2   41       1   83807.86              1   \n",
       "2          502    France       2   42       8  159660.80              3   \n",
       "3          699    France       2   39       1       0.00              2   \n",
       "4          850     Spain       2   43       2  125510.82              1   \n",
       "\n",
       "   HasCrCard  IsActiveMember  EstimatedSalary  Exited  \n",
       "0          1               1        101348.88       1  \n",
       "1          0               1        112542.58       0  \n",
       "2          1               0        113931.57       1  \n",
       "3          0               0         93826.63       0  \n",
       "4          1               1         79084.10       0  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>CreditScore</th>\n",
       "      <th>Geography</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure</th>\n",
       "      <th>Balance</th>\n",
       "      <th>NumOfProducts</th>\n",
       "      <th>HasCrCard</th>\n",
       "      <th>IsActiveMember</th>\n",
       "      <th>EstimatedSalary</th>\n",
       "      <th>Exited</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>619</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>42</td>\n",
       "      <td>2</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>101348.88</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>608</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>41</td>\n",
       "      <td>1</td>\n",
       "      <td>83807.86</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>112542.58</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>502</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>42</td>\n",
       "      <td>8</td>\n",
       "      <td>159660.80</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113931.57</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>699</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>39</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>93826.63</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>850</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>43</td>\n",
       "      <td>2</td>\n",
       "      <td>125510.82</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>79084.10</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   CreditScore  Geography  Gender  Age  Tenure    Balance  NumOfProducts  \\\n",
       "0          619          1       2   42       2       0.00              1   \n",
       "1          608          2       2   41       1   83807.86              1   \n",
       "2          502          1       2   42       8  159660.80              3   \n",
       "3          699          1       2   39       1       0.00              2   \n",
       "4          850          2       2   43       2  125510.82              1   \n",
       "\n",
       "   HasCrCard  IsActiveMember  EstimatedSalary  Exited  \n",
       "0          1               1        101348.88       1  \n",
       "1          0               1        112542.58       0  \n",
       "2          1               0        113931.57       1  \n",
       "3          0               0         93826.63       0  \n",
       "4          1               1         79084.10       0  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mapping Geography to Integer\n",
    "geography = {'France': 1,'Spain': 2,'Germany':3 } \n",
    "data.Geography = [geography[item] for item in data.Geography] \n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>Geography</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Gender</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.127134</td>\n",
       "      <td>0.131124</td>\n",
       "      <td>0.278116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.203450</td>\n",
       "      <td>0.212121</td>\n",
       "      <td>0.375524</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Geography         1         2         3\n",
       "Gender                                 \n",
       "1          0.127134  0.131124  0.278116\n",
       "2          0.203450  0.212121  0.375524"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Creating Pivot table for Data Analysis\n",
    "pivot = data.pivot_table(\"Exited\", index = \"Gender\", columns = \"Geography\")\n",
    "pivot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### From above table, we can detect \n",
    "1. More females have exited bank as compared to Males\n",
    "2. Higher rate of customers leaving in Germany"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build Machine Learning Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Steps for Preparing Data\n",
    "1. Random Shuffling of Data to avoid any Bias in Dataset\n",
    "2. Split Dataset into a-Features b-Target\n",
    "3. Split Data into Train and Test dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1.Shuffling dataset\n",
    "data = data.reindex(np.random.permutation(data.index))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 2.Split Dataset into a-Features b-Target\n",
    "feature = data.drop(\"Exited\", axis=1)\n",
    "target = data[\"Exited\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>CreditScore</th>\n",
       "      <th>Geography</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure</th>\n",
       "      <th>Balance</th>\n",
       "      <th>NumOfProducts</th>\n",
       "      <th>HasCrCard</th>\n",
       "      <th>IsActiveMember</th>\n",
       "      <th>EstimatedSalary</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8516</th>\n",
       "      <td>500</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>34</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>12833.96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8202</th>\n",
       "      <td>589</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>27</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>144181.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9668</th>\n",
       "      <td>720</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>34</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>77047.78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4957</th>\n",
       "      <td>752</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>39</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>188187.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6160</th>\n",
       "      <td>632</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>21</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>84008.66</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      CreditScore  Geography  Gender  Age  Tenure  Balance  NumOfProducts  \\\n",
       "8516          500          1       2   34       4      0.0              2   \n",
       "8202          589          2       2   27       4      0.0              2   \n",
       "9668          720          2       1   34       3      0.0              2   \n",
       "4957          752          1       1   39       3      0.0              1   \n",
       "6160          632          1       2   21       1      0.0              2   \n",
       "\n",
       "      HasCrCard  IsActiveMember  EstimatedSalary  \n",
       "8516          1               0         12833.96  \n",
       "8202          1               0        144181.48  \n",
       "9668          1               1         77047.78  \n",
       "4957          1               0        188187.05  \n",
       "6160          1               0         84008.66  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature.head()\n",
    "#target.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing Machine Learning Libraries for Test/Train Split of Data\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Splitting Features from \n",
    "X_train, X_test, y_train, y_test = train_test_split(feature,target, test_size = 0.2, random_state=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### This is \"Supervised\" - \"Classification\" Problem - since data is labelled and can be classified into two distict classes - Exited (0/1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predicting with Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import Library\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import accuracy_score, confusion_matrix\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression : 80.55%\n"
     ]
    }
   ],
   "source": [
    "clf = LogisticRegression()\n",
    "clf.fit(X_train, y_train)\n",
    "pred = clf.predict(X_test)\n",
    "print(\"Logistic Regression : {}%\".format(accuracy_score(pred,y_test)*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predicting with Decision Tree "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Most Machine Learning Algorithms are classified into 2 categories:\n",
    "1. Distance Based - KNN, SVM - work efficiently with feature scaling\n",
    "2. Density Based - Naive Bayes, Decision Tree, Random Forest ensemble methods - are not much affected by Feature Scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import Library\n",
    "from sklearn import tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree : 78.95%\n"
     ]
    }
   ],
   "source": [
    "clf = tree.DecisionTreeClassifier()\n",
    "clf.fit(X_train, y_train)\n",
    "pred = clf.predict(X_test)\n",
    "accuracy_score(pred, y_test)\n",
    "print(\"Decision Tree : {}%\".format(accuracy_score(pred,y_test)*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predicting with Gaussian Naive Bayes Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import Library\n",
    "from sklearn.naive_bayes import GaussianNB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gaussian Naive Bayes : 80.75%\n"
     ]
    }
   ],
   "source": [
    "clf = GaussianNB()\n",
    "clf.fit(X_train, y_train)\n",
    "pred = clf.predict(X_test)\n",
    "print(\"Gaussian Naive Bayes : {}%\".format(accuracy_score(pred,y_test)*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predicting with Random Forest Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import Library\n",
    "from sklearn.ensemble import RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest Model : 87.05%\n"
     ]
    }
   ],
   "source": [
    "clf = RandomForestClassifier(n_estimators = 200, random_state=200)\n",
    "clf.fit(X_train, y_train)\n",
    "pred = clf.predict(X_test)\n",
    "print(\"Random Forest Model : {:.2f}%\".format(accuracy_score(pred,y_test)*100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1551,   72],\n",
       "       [ 187,  190]], dtype=int64)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "confusion_matrix(y_test, pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Scaling (Normalization)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### Some algorithm need Feature Scaling for better accuracy - For e.g. SVM, K-Nearest Neighbour, K-Means Clustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import Preprocessing Libraries\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.preprocessing import MinMaxScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Feature Scaling using sklearn Preprocessing Module\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler = StandardScaler()\n",
    "X_train = scaler.fit_transform(X_train)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-1.11172518, -0.89555975, -0.91324755, ...,  0.64842466,\n",
       "        -1.03097965, -0.2520836 ],\n",
       "       [-0.53090211,  0.30794739, -0.91324755, ...,  0.64842466,\n",
       "        -1.03097965,  0.19314186],\n",
       "       [-2.08667819, -0.89555975,  1.09499335, ..., -1.5421992 ,\n",
       "        -1.03097965,  1.30660683],\n",
       "       ...,\n",
       "       [ 0.56851299,  1.51145453,  1.09499335, ...,  0.64842466,\n",
       "         0.96995125, -0.01778769],\n",
       "       [-1.0391223 ,  1.51145453,  1.09499335, ..., -1.5421992 ,\n",
       "        -1.03097965, -0.0499569 ],\n",
       "       [ 2.07242988,  1.51145453,  1.09499335, ...,  0.64842466,\n",
       "        -1.03097965,  0.1954521 ]])"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#X_train\n",
    "X_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predicting with Support Vector Machine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import Library\n",
    "from sklearn.svm import SVC  #Support Vector Classification module  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto', kernel='linear',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "svclassifier = SVC(kernel='linear')  \n",
    "svclassifier.fit(X_train, y_train)  #Train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Making Prediction\n",
    "y_pred = svclassifier.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Support Vector Classifier : 87.05%\n"
     ]
    }
   ],
   "source": [
    "print(\"Support Vector Classifier : {:.2f}%\".format(accuracy_score(pred,y_test)*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predicting with K-Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import Library\n",
    "from sklearn.neighbors import KNeighborsClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "           metric_params=None, n_jobs=1, n_neighbors=5, p=2,\n",
       "           weights='uniform')"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knnclassifier = KNeighborsClassifier()\n",
    "knnclassifier.fit(X_train, y_train)  #Train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Making Prediction\n",
    "y_pred = knnclassifier.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K-Nearest Neighbour Classifier : 87.05%\n"
     ]
    }
   ],
   "source": [
    "print(\"K-Nearest Neighbour Classifier : {:.2f}%\".format(accuracy_score(pred,y_test)*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluating Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#1 Accuracy Score\n",
    "#2 Confusion Matrix\n",
    "#3 Classification Report\n",
    "#4 Area under Curve    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'             precision    recall  f1-score   support\\n\\n          0       0.87      0.94      0.90      1623\\n          1       0.59      0.38      0.47       377\\n\\navg / total       0.82      0.83      0.82      2000\\n'"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "classification_report(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Precision: This a measure of a classifier's exactness, i.e., the ability of a classifier to give correct labels.\n",
    "\n",
    "Recall: A measure of a classifiers completeness, i.e., the ability of a classifier to find all samples that belong a certain class.\n",
    "\n",
    "F1 Score (or F-score): A weighted average of precision and recall."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Testing Model\n",
    "Select the best algorithm and test with New Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest Model : 86.95%\n"
     ]
    }
   ],
   "source": [
    "#Random Forest Classifier - Accuracy Score = 87%\n",
    "model = RandomForestClassifier(n_estimators = 200, random_state=200)\n",
    "model.fit(X_train, y_train)\n",
    "pred = model.predict(X_test)\n",
    "print(\"Random Forest Model : {:.2f}%\".format(accuracy_score(pred,y_test)*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Get Input values of All Features from user and predict if Customer will exit or not"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Enter Credit Score :\n",
      "619\n",
      "Enter Geography ('France': 1,'Spain': 2,'Germany':3 ) :\n",
      "1\n",
      "Enter Gender ('Male': 1,'Female': 2) :\n",
      "2\n",
      "Enter Age :\n",
      "42\n",
      "Enter Tenure :\n",
      "2\n",
      "Enter Current Balance :\n",
      "0\n",
      "Enter Number of Products :\n",
      "1\n",
      "Has Credit Card ? ('Yes': 1,'No': 0) :\n",
      "1\n",
      "Is Active Member ? ('Yes': 1,'No': 0) :\n",
      "1\n",
      "Enter Estimated Salary :\n",
      "101348\n"
     ]
    }
   ],
   "source": [
    "print(\"Enter Credit Score :\")\n",
    "CreditScore = int(input())\n",
    "print(\"Enter Geography ('France': 1,'Spain': 2,'Germany':3 ) :\")\n",
    "Geography = int(input())\n",
    "print(\"Enter Gender ('Male': 1,'Female': 2) :\")\n",
    "Gender = int(input())\n",
    "print(\"Enter Age :\")\n",
    "Age = int(input())\n",
    "print(\"Enter Tenure :\")\n",
    "Tenure = int(input())\n",
    "print(\"Enter Current Balance :\")\n",
    "Balance = float(input())\n",
    "print(\"Enter Number of Products :\")\n",
    "NumOfProducts = int(input())\n",
    "print(\"Has Credit Card ? ('Yes': 1,'No': 0) :\")\n",
    "HasCrCard = int(input())\n",
    "print(\"Is Active Member ? ('Yes': 1,'No': 0) :\")\n",
    "IsActiveMember = int(input())\n",
    "print(\"Enter Estimated Salary :\")\n",
    "EstimatedSalary = float(input())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_user = np.array([CreditScore,Geography,Gender,Age, Tenure, Balance, NumOfProducts, HasCrCard, IsActiveMember, EstimatedSalary])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([6.19000e+02, 1.00000e+00, 2.00000e+00, 4.20000e+01, 2.00000e+00,\n",
       "       0.00000e+00, 1.00000e+00, 1.00000e+00, 1.00000e+00, 1.01348e+05])"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_user"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = model.predict([X_user])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "index = y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Customer might soon exit the Bank, Please take Action !!\n"
     ]
    }
   ],
   "source": [
    "if index == 1:\n",
    "    print(\"Customer will not exit the Bank\")\n",
    "else:\n",
    "    print(\"Customer might soon exit the Bank, Please take Action !!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Derived Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have accuracy of maximum 90% with 1000 rows. This accuracy can be impoved by collection more data and retraining Algorithm iteratively. Algorithm will continue learning from newer Data and better domain knowledge and get optimized to increase Accuracy Score.\n",
    "\n",
    "Credit Score, Product, Credit Card or Salary has lower impact.\n",
    "Age is important factor - older customers are exiting more than younger ones. \n",
    "Females exit ratio is more than Males. \n",
    "Germany is least performing Country. \n",
    "Customers with low Tenure or high Tenure and leaving more than ones in mid."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}