LoanPrediction_ANN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "qOVWFS4166uB"
   },
   "outputs": [],
   "source": [
    "import numpy as np \n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score, classification_report\n",
    "%matplotlib inline\n",
    "\n",
    "np.set_printoptions(formatter={'float_kind':'{:f}'.format})\n",
    "\n",
    "sns.set(rc={'figure.figsize':(12,10)})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 383
    },
    "id": "jYogUN0893ME",
    "outputId": "6c8de75c-bb36-4925-d0d1-831c693376f9"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1006, 14)\n"
     ]
    },
    {
     "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>Type of Account</th>\n",
       "      <th>Account History</th>\n",
       "      <th>Reason for the Loan</th>\n",
       "      <th>Loan Amount</th>\n",
       "      <th>Account Savings</th>\n",
       "      <th>Employment History</th>\n",
       "      <th>Individual Stauts</th>\n",
       "      <th>Other Loans</th>\n",
       "      <th>Security / Collateral</th>\n",
       "      <th>Age</th>\n",
       "      <th>Residence Status</th>\n",
       "      <th>Job</th>\n",
       "      <th>Completed Other loan?</th>\n",
       "      <th>Good Loan</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>No Data</td>\n",
       "      <td>critical</td>\n",
       "      <td>Learning / Edu purposes</td>\n",
       "      <td>2116.0</td>\n",
       "      <td>0-200</td>\n",
       "      <td>5-7 Years</td>\n",
       "      <td>Male</td>\n",
       "      <td>No</td>\n",
       "      <td>Property - Real Estate</td>\n",
       "      <td>47.0</td>\n",
       "      <td>House Owner</td>\n",
       "      <td>Services</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>No Data</td>\n",
       "      <td>critical</td>\n",
       "      <td>TV</td>\n",
       "      <td>2438.0</td>\n",
       "      <td>No Data</td>\n",
       "      <td>7+ Years</td>\n",
       "      <td>Male</td>\n",
       "      <td>No</td>\n",
       "      <td>Savings Account</td>\n",
       "      <td>52.0</td>\n",
       "      <td>House Owner</td>\n",
       "      <td>Skilled Labor</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>No Data</td>\n",
       "      <td>critical</td>\n",
       "      <td>Buying a New Car</td>\n",
       "      <td>2144.0</td>\n",
       "      <td>0-200</td>\n",
       "      <td>2-5 Years</td>\n",
       "      <td>Male</td>\n",
       "      <td>No</td>\n",
       "      <td>Vehicle</td>\n",
       "      <td>48.0</td>\n",
       "      <td>House Owner</td>\n",
       "      <td>Skilled Labor</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>No Data</td>\n",
       "      <td>critical</td>\n",
       "      <td>Home furniture</td>\n",
       "      <td>2073.0</td>\n",
       "      <td>No Data</td>\n",
       "      <td>2-5 Years</td>\n",
       "      <td>Married</td>\n",
       "      <td>No</td>\n",
       "      <td>Vehicle</td>\n",
       "      <td>25.0</td>\n",
       "      <td>House Owner</td>\n",
       "      <td>Skilled Labor</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>No Data</td>\n",
       "      <td>critical</td>\n",
       "      <td>Support for Business</td>\n",
       "      <td>1276.0</td>\n",
       "      <td>No Data</td>\n",
       "      <td>7+ Years</td>\n",
       "      <td>Male</td>\n",
       "      <td>No</td>\n",
       "      <td>No Security</td>\n",
       "      <td>59.0</td>\n",
       "      <td>Renting</td>\n",
       "      <td>Services</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Type of Account Account History      Reason for the Loan  Loan Amount  \\\n",
       "0         No Data        critical  Learning / Edu purposes       2116.0   \n",
       "1         No Data        critical                       TV       2438.0   \n",
       "2         No Data        critical         Buying a New Car       2144.0   \n",
       "3         No Data        critical           Home furniture       2073.0   \n",
       "4         No Data        critical     Support for Business       1276.0   \n",
       "\n",
       "  Account Savings Employment History Individual Stauts Other Loans  \\\n",
       "0           0-200          5-7 Years              Male          No   \n",
       "1         No Data           7+ Years              Male          No   \n",
       "2           0-200          2-5 Years              Male          No   \n",
       "3         No Data          2-5 Years           Married          No   \n",
       "4         No Data           7+ Years              Male          No   \n",
       "\n",
       "    Security / Collateral   Age Residence Status            Job  \\\n",
       "0  Property - Real Estate  47.0      House Owner       Services   \n",
       "1         Savings Account  52.0      House Owner  Skilled Labor   \n",
       "2                 Vehicle  48.0      House Owner  Skilled Labor   \n",
       "3                 Vehicle  25.0      House Owner  Skilled Labor   \n",
       "4             No Security  59.0          Renting       Services   \n",
       "\n",
       "  Completed Other loan? Good Loan  \n",
       "0                    No        No  \n",
       "1                    No        No  \n",
       "2                    No        No  \n",
       "3                   Yes        No  \n",
       "4                    No        No  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('https://raw.githubusercontent.com/Pitsillides91/Python-Tutorials/master/Introduction%20to%20ML%20-%20Logistic%20Regression%20Example/Logistic%20Regression%20Dummy%20Data%20v3.csv')\n",
    "print(df.shape)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "PYoypBeT980-"
   },
   "outputs": [],
   "source": [
    "### Exploratory data analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "JHKFE9d7-X03",
    "outputId": "e8fb5137-e899-497f-80db-a7991d42c735"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Type of Account          2\n",
       "Account History          2\n",
       "Reason for the Loan      3\n",
       "Loan Amount              3\n",
       "Account Savings          4\n",
       "Employment History       5\n",
       "Individual Stauts        5\n",
       "Other Loans              5\n",
       "Security / Collateral    5\n",
       "Age                      4\n",
       "Residence Status         4\n",
       "Job                      4\n",
       "Completed Other loan?    4\n",
       "Good Loan                5\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 333
    },
    "id": "8c0-A-5Y-gKK",
    "outputId": "b445331f-d1ce-4d74-c8b9-77f4197875e0"
   },
   "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>Type of Account</th>\n",
       "      <th>Account History</th>\n",
       "      <th>Reason for the Loan</th>\n",
       "      <th>Loan Amount</th>\n",
       "      <th>Account Savings</th>\n",
       "      <th>Employment History</th>\n",
       "      <th>Individual Stauts</th>\n",
       "      <th>Other Loans</th>\n",
       "      <th>Security / Collateral</th>\n",
       "      <th>Age</th>\n",
       "      <th>Residence Status</th>\n",
       "      <th>Job</th>\n",
       "      <th>Completed Other loan?</th>\n",
       "      <th>Good Loan</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>686</th>\n",
       "      <td>Type C</td>\n",
       "      <td>Good</td>\n",
       "      <td>Buying a New Car</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>721</th>\n",
       "      <td>Type C</td>\n",
       "      <td>critical</td>\n",
       "      <td>Home furniture</td>\n",
       "      <td>5387.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>737</th>\n",
       "      <td>Type C</td>\n",
       "      <td>Good</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1004</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>43.0</td>\n",
       "      <td>House Owner</td>\n",
       "      <td>Services</td>\n",
       "      <td>No</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1005</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1765.0</td>\n",
       "      <td>0-201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Type of Account Account History Reason for the Loan  Loan Amount  \\\n",
       "686           Type C            Good    Buying a New Car          NaN   \n",
       "721           Type C        critical      Home furniture       5387.0   \n",
       "737           Type C            Good                 NaN          NaN   \n",
       "1004             NaN             NaN                 NaN          NaN   \n",
       "1005             NaN             NaN                 NaN       1765.0   \n",
       "\n",
       "     Account Savings Employment History Individual Stauts Other Loans  \\\n",
       "686              NaN                NaN               NaN         NaN   \n",
       "721              NaN                NaN               NaN         NaN   \n",
       "737              NaN                NaN               NaN         NaN   \n",
       "1004             NaN                NaN               NaN         NaN   \n",
       "1005           0-201                NaN               NaN         NaN   \n",
       "\n",
       "     Security / Collateral   Age Residence Status       Job  \\\n",
       "686                    NaN   NaN              NaN       NaN   \n",
       "721                    NaN   NaN              NaN       NaN   \n",
       "737                    NaN   NaN              NaN       NaN   \n",
       "1004                   NaN  43.0      House Owner  Services   \n",
       "1005                   NaN   NaN              NaN       NaN   \n",
       "\n",
       "     Completed Other loan? Good Loan  \n",
       "686                    NaN       NaN  \n",
       "721                    NaN       NaN  \n",
       "737                    NaN       NaN  \n",
       "1004                    No       NaN  \n",
       "1005                   NaN       NaN  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df['Employment History'].isnull()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "qa8Ki_1s-ine",
    "outputId": "9da60cf8-ac40-4057-82c0-816a6162d838"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Type of Account          0\n",
       "Account History          0\n",
       "Reason for the Loan      0\n",
       "Loan Amount              0\n",
       "Account Savings          0\n",
       "Employment History       0\n",
       "Individual Stauts        0\n",
       "Other Loans              0\n",
       "Security / Collateral    0\n",
       "Age                      0\n",
       "Residence Status         0\n",
       "Job                      0\n",
       "Completed Other loan?    0\n",
       "Good Loan                0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df.dropna(subset=['Employment History'], axis = 0)\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1001, 14)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape  # only 5 rows are deleted. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "izYUrTcw-mUx",
    "outputId": "96c225f5-7351-40a6-9b00-503f1c0c3a1d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 1001 entries, 0 to 1003\n",
      "Data columns (total 14 columns):\n",
      " #   Column                 Non-Null Count  Dtype  \n",
      "---  ------                 --------------  -----  \n",
      " 0   Type of Account        1001 non-null   object \n",
      " 1   Account History        1001 non-null   object \n",
      " 2   Reason for the Loan    1001 non-null   object \n",
      " 3   Loan Amount            1001 non-null   float64\n",
      " 4   Account Savings        1001 non-null   object \n",
      " 5   Employment History     1001 non-null   object \n",
      " 6   Individual Stauts      1001 non-null   object \n",
      " 7   Other Loans            1001 non-null   object \n",
      " 8   Security / Collateral  1001 non-null   object \n",
      " 9   Age                    1001 non-null   float64\n",
      " 10  Residence Status       1001 non-null   object \n",
      " 11  Job                    1001 non-null   object \n",
      " 12  Completed Other loan?  1001 non-null   object \n",
      " 13  Good Loan              1001 non-null   object \n",
      "dtypes: float64(2), object(12)\n",
      "memory usage: 117.3+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "V2Xq2l2v-o60",
    "outputId": "a258913e-42fc-47b6-c2db-dab82eca1072"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of unique values of feature Type of Account is ['No Data' 'Type A' 'Type B' 'Type C']\n",
      "The number of unique values of feature Account History is ['critical' 'Average (known delays)' 'Good']\n",
      "The number of unique values of feature Account Savings is ['0-200' 'No Data' '200-500' '1000+' '500-1000']\n",
      "The number of unique values of feature Employment History is ['5-7 Years' '7+ Years' '2-5 Years' '0-2 Years' 'Unemployed']\n",
      "The number of unique values of feature Individual Stauts is ['Male' 'Married' 'Female' 'Divorced' 'Female_']\n",
      "The number of unique values of feature Other Loans is ['No' 'Yes']\n",
      "The number of unique values of feature Security / Collateral is ['Property - Real Estate' 'Savings Account' 'Vehicle' 'No Security']\n",
      "The number of unique values of feature Residence Status is ['House Owner' 'Renting' 'Free']\n",
      "The number of unique values of feature Job is ['Services' 'Skilled Labor' 'Professional / Management' 'Not Employed']\n",
      "The number of unique values of feature Completed Other loan? is ['No' 'Yes']\n",
      "The number of unique values of feature Good Loan is ['No' 'Yes']\n"
     ]
    }
   ],
   "source": [
    "## check for features with unique elements less than 10\n",
    "\n",
    "for col in df:\n",
    "    if df[col].nunique() < 10:\n",
    "        print('The number of unique values of feature {} is {}'.format(col,df[col].unique()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "zHggPaOD-r5p",
    "outputId": "b3c629bf-ef11-4b85-d3e0-8a77ace41bdb"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of unique values of feature Loan Amount is 911\n",
      "The number of unique values of feature Age is 57\n"
     ]
    }
   ],
   "source": [
    "for col in df:\n",
    "    if df[col].nunique() > 10:\n",
    "        print('The number of unique values of feature {} is {}'.format(col,df[col].nunique()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "m_ZvklwZ-wPi"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x1d1e6496370>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## visualize loan and age\n",
    "\n",
    "sns.pairplot(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "id": "LcTiuhqJ-yGK"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x1d1e6631730>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 427.925x360 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## remove the outlier to get clear visualization\n",
    "\n",
    "df = df[df['Age'] < 100]\n",
    "\n",
    "sns.pairplot(df, hue = 'Good Loan')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "l592VKi4LRvm"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1d1e65216a0>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtUAAAJPCAYAAACzTihOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3df5DU9WH/8dfCIToS44/cgSGUUZuEDv1GbDoqVo8ktgdCz+KhEyOGqDFjrMHEWi3yKxG1okNLNS2t0zhOa00j5YdQQk872jFVtCqZyJCxjU2EBuIcBwJ60Tvlbr9/5Jvr16op5H3r3iWPx1/sZ/duXwMz63PW991WqtVqNQAAwM9tWL0HAADAUCeqAQCgkKgGAIBCohoAAAqJagAAKCSqAQCgkKgGAIBCDfUeMFD27v1x+vr8ym0AAGpj2LBKjjnmyLe97xcmqvv6qqIaAIC6cPwDAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoFBDvQf8InnPUYfn8JEj6j0DGOS6e97IKy9313sGAANIVA+gw0eOyEXX31fvGcAg9/XbZ+eViGqAXySOfwAAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQKGGWn7zT3/603nppZfS0PCTp1myZEl+/OMf59Zbb01PT0/OOeecXHPNNUmS5557LgsWLMiPf/zj/OZv/mZuvPHG/q8DAIDBrGbVWq1Ws23btvzLv/xLfxx3d3dn2rRpuffee3P88cfniiuuyKOPPpopU6bkuuuuy80335xJkyZl/vz5WblyZS666KJazQMAgAFTs+MfP/jBD5Ikl112Wc4999z83d/9XbZs2ZLx48dn3LhxaWhoSGtra9rb27Nz5850d3dn0qRJSZK2tra0t7fXahoAAAyomr1T/fLLL2fy5MlZtGhR3njjjcyZMyeXX355Ghsb+x/T1NSUjo6O7Nq1603XGxsb09HRcUjPd9xxowZsO0CtNTa+p94TABhANYvqU045Jaecckr/7fPPPz933nlnPvrRj/Zfq1arqVQq6evrS6VSecv1Q7FnT1f6+qrlwwv4jyRwsDo7X6n3BAAO0bBhlXd8I7dmxz+eeeaZPPHEE/23q9Vqxo4dm87Ozv5rnZ2daWpqypgxY950fffu3WlqaqrVNAAAGFA1i+pXXnklt99+e3p6etLV1ZW1a9fmD/7gD/LCCy9k+/bt6e3tzYYNG9Lc3JyxY8dm5MiR2bx5c5Jk3bp1aW5urtU0AAAYUDU7/vHxj388zz77bGbOnJm+vr5cdNFFOeWUU7J06dLMnTs3PT09mTJlSqZNm5YkWbZsWRYuXJiurq5MnDgxc+bMqdU0AAAYUJVqtVrfg8gDZLCcqb7o+vvqugEY/L5++2xnqgGGoLqcqQYAgF8WohoAAAqJagAAKCSqAQCgkKgGAIBCohoAAAqJagAAKCSqAQCgkKgGAIBCohoAAAqJagAAKCSqAQCgkKgGAIBCohoAAAqJagAAKCSqAQCgkKgGAIBCohoAAAqJagAAKCSqAQCgkKgGAIBCohoAAAqJagAAKCSqAQCgkKgGAIBCohoAAAqJagAAKCSqAQCgkKgGAIBCDfUeAMAvr2Pee1gaDhtZ7xnAEHDg9Z7s3f96vWe8I1ENQN00HDYym2+/vN4zgCHgo9d/LcngjWrHPwAAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAAChU86i+7bbbMm/evCTJpk2b0trampaWlixfvrz/Mc8991za2toyderULFiwIAcOHKj1LAAAGDA1jeonnngia9euTZJ0d3dn/vz5WbFiRTZu3JitW7fm0UcfTZJcd911Wbx4cR588MFUq9WsXLmylrMAAGBA1Syq9+3bl+XLl+fzn/98kmTLli0ZP358xo0bl4aGhrS2tqa9vT07d+5Md3d3Jk2alCRpa2tLe3t7rWYBAMCAq1lUL168ONdcc02OOuqoJMmuXbvS2NjYf39TU1M6Ojrecr2xsTEdHR21mgUAAAOuoRbf9B/+4R9y/PHHZ/LkyVmzZk2SpK+vL5VKpf8x1Wo1lUrlHa8fquOOG1U+HOBd0tj4nnpPABhyBvNrZ02ieuPGjens7Mzv/d7vZf/+/Xn11Vezc+fODB8+vP8xnZ2daWpqypgxY9LZ2dl/fffu3Wlqajrk59yzpyt9fdUB2f/zGsz/0MDg0tn5Sr0nDApeN4FDUe/XzmHDKu/4Rm5Novqee+7p//OaNWvy1FNP5cYbb0xLS0u2b9+eD3zgA9mwYUNmzZqVsWPHZuTIkdm8eXM++tGPZt26dWlubq7FLAAAqImaRPXbGTlyZJYuXZq5c+emp6cnU6ZMybRp05Iky5Yty8KFC9PV1ZWJEydmzpw579YsAAAoVvOobmtrS1tbW5Jk8uTJWb9+/VseM2HChKxatarWUwAAoCZ8oiIAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFahrVd9xxR6ZPn54ZM2bknnvuSZJs2rQpra2taWlpyfLly/sf+9xzz6WtrS1Tp07NggULcuDAgVpOAwCAAVOzqH7qqafy5JNPZv369Vm9enXuvffe/Pu//3vmz5+fFStWZOPGjdm6dWseffTRJMl1112XxYsX58EHH0y1Ws3KlStrNQ0AAAZUzaL61FNPzd/+7d+moaEhe/bsSW9vb15++eWMHz8+48aNS0NDQ1pbW9Pe3p6dO3emu7s7kyZNSpK0tbWlvb29VtMAAGBA1fT4x4gRI3LnnXdmxowZmTx5cnbt2pXGxsb++5uamtLR0fGW642Njeno6KjlNAAAGDANtX6Cq6++Op/73Ofy+c9/Ptu2bUulUum/r1qtplKppK+v722vH4rjjhs1YJsBaq2x8T31ngAw5Azm186aRfX3v//9vP766/m1X/u1HHHEEWlpaUl7e3uGDx/e/5jOzs40NTVlzJgx6ezs7L++e/fuNDU1HdLz7dnTlb6+6oDt/3kM5n9oYHDp7Hyl3hMGBa+bwKGo92vnsGGVd3wjt2bHP3bs2JGFCxfm9ddfz+uvv56HH344F154YV544YVs3749vb292bBhQ5qbmzN27NiMHDkymzdvTpKsW7cuzc3NtZoGAAADqmbvVE+ZMiVbtmzJzJkzM3z48LS0tGTGjBk59thjM3fu3PT09GTKlCmZNm1akmTZsmVZuHBhurq6MnHixMyZM6dW0wAAYEDV9Ez13LlzM3fu3Dddmzx5ctavX/+Wx06YMCGrVq2q5RwAAKiJgzr+8Xa/ieM///M/B3wMAAAMRT8zqvft25d9+/blc5/7XPbv399/e/fu3fnCF77wbm0EAIBB7Wce/7j22mvz+OOPJ0lOO+20//6ihoZMnTq1tssAAGCI+JlRfffddydJbrjhhtx6663vyiAAABhqDuoHFW+99dbs3Lkz+/fvT7X6378LeuLEiTUbBgAAQ8VBRfWdd96Zu+++O8cdd1z/tUqlkocffrhmwwAAYKg4qKh+4IEH8tBDD2X06NG13gMAAEPOQf1KveOPP15QAwDAOziod6onT56c22+/PWeffXYOP/zw/uvOVAMAwEFG9Zo1a5Ik7e3t/decqQYAgJ84qKh+5JFHar0DAACGrIOK6nvuuedtr1966aUDOgYAAIaig4rq733ve/1/fv311/P0009n8uTJNRsFAABDyUF/+Mv/r6OjIwsWLKjJIAAAGGoO6lfq/U+jR4/Ozp07B3oLAAAMSYd8prparWbr1q1v+nRFAAD4ZXbIZ6qTn3wYzPXXX1+TQQAAMNQc0pnqnTt35sCBAxk/fnxNRwEAwFByUFG9ffv2/P7v/3527dqVvr6+HHPMMbnrrrty0kkn1XofAAAMegf1g4pLlizJ5ZdfnqeffjqbN2/OlVdemRtvvLHW2wAAYEg4qKjes2dPzjvvvP7bs2bNyt69e2s2CgAAhpKDiure3t7s27ev//ZLL71Us0EAADDUHNSZ6osvvjif/OQnc84556RSqWTjxo35zGc+U+ttAAAwJBzUO9VTpkxJkrzxxhv5/ve/n46OjvzO7/xOTYcBAMBQcVDvVM+bNy+zZ8/OnDlz0tPTk7//+7/P/Pnz89d//de13gcAAIPeQb1TvXfv3syZMydJMnLkyFxyySXp7Oys6TAAABgqDvoHFTs6Ovpv7969O9VqtWajAABgKDmo4x+XXHJJZs6cmbPOOiuVSiWbNm3yMeUAAPD/HFRUn3/++fn1X//1PPnkkxk+fHg++9nP5kMf+lCttwEAwJBwUFGdJBMmTMiECRNquQUAAIakgzpTDQAAvDNRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhUQ1AAAUEtUAAFBIVAMAQCFRDQAAhWoa1X/+53+eGTNmZMaMGbn99tuTJJs2bUpra2taWlqyfPny/sc+99xzaWtry9SpU7NgwYIcOHCgltMAAGDA1CyqN23alMceeyxr167NAw88kO9+97vZsGFD5s+fnxUrVmTjxo3ZunVrHn300STJddddl8WLF+fBBx9MtVrNypUrazUNAAAGVM2iurGxMfPmzcthhx2WESNG5KSTTsq2bdsyfvz4jBs3Lg0NDWltbU17e3t27tyZ7u7uTJo0KUnS1taW9vb2Wk0DAIAB1VCrb/zBD36w/8/btm3LP/3TP+Xiiy9OY2Nj//WmpqZ0dHRk165db7re2NiYjo6OQ3q+444bVT4a4F3S2Pieek8AGHIG82tnzaL6p55//vlcccUVuf766zN8+PBs27at/75qtZpKpZK+vr5UKpW3XD8Ue/Z0pa+vOlCzfy6D+R8aGFw6O1+p94RBwesmcCjq/do5bFjlHd/IrekPKm7evDmXXHJJrr322px33nkZM2ZMOjs7++/v7OxMU1PTW67v3r07TU1NtZwGAAADpmZR/eKLL+aqq67KsmXLMmPGjCTJySefnBdeeCHbt29Pb29vNmzYkObm5owdOzYjR47M5s2bkyTr1q1Lc3NzraYBAMCAqtnxj7vvvjs9PT1ZunRp/7ULL7wwS5cuzdy5c9PT05MpU6Zk2rRpSZJly5Zl4cKF6erqysSJEzNnzpxaTQMAgAFVs6heuHBhFi5c+Lb3rV+//i3XJkyYkFWrVtVqDgAA1IxPVAQAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoJKoBAKCQqAYAgEKiGgAAColqAAAoVNOo7urqyu/+7u9mx44dSZJNmzaltbU1LS0tWb58ef/jnnvuubS1tWXq1KlZsGBBDhw4UMtZAAAwoGoW1c8++2w+9alPZdu2bUmS7u7uzJ8/PytWrMjGjRuzdevWPProo0mS6667LosXL86DDz6YarWalStX1moWAAAMuJpF9cqVK/PlL385TU1NSZItW7Zk/PjxGTduXBoaGtLa2pr29vbs3Lkz3d3dmTRpUpKkra0t7e3ttZoFAAADrqFW3/iWW2550+1du3alsbGx/3ZTU1M6Ojrecr2xsTEdHR21mgUAAAOuZlH9P/X19aVSqfTfrlarqVQq73j9UB133KgB2QnwbmhsfE+9JwAMOYP5tfNdi+oxY8aks7Oz/3ZnZ2eamprecn337t39R0YOxZ49Xenrqw7I1p/XYP6HBgaXzs5X6j1hUPC6CRyKer92DhtWecc3ct+1X6l38skn54UXXsj27dvT29ubDRs2pLm5OWPHjs3IkSOzefPmJMm6devS3Nz8bs0CAIBi79o71SNHjszSpUszd+7c9PT0ZMqUKZk2bVqSZNmyZVm4cGG6uroyceLEzJkz592aBQAAxWoe1Y888kj/nydPnpz169e/5TETJkzIqlWraj0FAABqwicqAgBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUEhUAwBAIVENAACFRDUAABQS1QAAUGhQRfU//uM/Zvr06Wlpacl9991X7zkAAHBQGuo94Kc6OjqyfPnyrFmzJocddlguvPDCnHbaafnVX/3Vek8DAICfadBE9aZNm3L66afn6KOPTpJMnTo17e3t+cIXvnBQXz9sWKWW8w7a+445st4TgCFgsLxmDQaHHXVcvScAQ0S9Xzt/1vMPmqjetWtXGhsb+283NTVly5YtB/31xwySmL3zhpn1ngAMAccdN6reEwaN//P52+o9ARgiBvNr56A5U93X15dK5b/rv1qtvuk2AAAMVoMmqseMGZPOzs7+252dnWlqaqrjIgAAODiDJqrPOOOMPPHEE3nppZfy2muv5aGHHkpzc3O9ZwEAwP9q0JypHj16dK655prMmTMnb7zxRs4///x85CMfqfcsAAD4X1Wq1Wq13iMAAGAoGzTHPwAAYKgS1QAAUEhUAwBAIVENAACFRDUMgB07duTDH/5wHn/88Tdd/8QnPpEdO3bUaRXA4LNkyZJcffXVb7r22GOP5eyzz05XV1edVkE5UQ0DZMSIEVm0aJH/KAD8DNdee222bt2ahx9+OEny6quv5itf+Ur++I//OKNGDd6PoIb/jaiGAdLU1JQzzjgjt91221vu+6u/+qtMnz49ra2tWbp0aXp7e+uwEKD+jjzyyNx888256aab8uqrr+bOO+/MJz7xiRxxxBH51Kc+lfPOOy+XXXZZfvjDHyZJ7rnnnpx77rmZOXNmFi9eXOf18M5ENQygefPm5bHHHnvTMZBvfetbeeSRR7J69eqsXbs227dvzze+8Y06rgSorzPOOCNnnnlmbrjhhjz++OOZO3duFi5cmD/5kz/J2rVrc+mll2bRokXp7e3NXXfdldWrV2fNmjV544030tHRUe/58LYGzScqwi+CUaNG5aabbsqiRYuyfv36JMmTTz6ZGTNm5IgjjkiSzJo1Kw888EBmz55dz6kAdTVv3rx87GMfy1/8xV/kxRdfzA9/+MNceeWV/fd3dXVl+PDhOeWUU3L++efn7LPPzqWXXprRo0fXcTW8M1ENA+zMM8980zGQvr6+tzzmwIED7/YsgEFl1KhROeqoozJ27Nh0dXXlAx/4QNatW5ck6e3tze7du5MkK1asyHe+851861vfyuWXX55ly5bl1FNPred0eFuOf0AN/PQYyK5du3L66afnm9/8Zrq7u3PgwIGsXr06p59+er0nAgwaJ554Yvbv359nnnkmSbJ69er84R/+YV566aVMnz49H/rQh/LFL34xv/Vbv5X/+I//qPNaeHveqYYa+OkxkM9+9rP52Mc+lpdffjmzZs3KgQMHcuaZZ+biiy+u90SAQeOwww7LHXfckVtuuSU9PT0ZNWpUbrvtthx77LH55Cc/mfPPPz9HHHFETjjhhMyaNavec+FtVarVarXeIwAAYChz/AMAAAqJagAAKCSqAQCgkKgGAIBCohoAAAqJaoBBYtWqVbngggsyffr0/PZv/3YuvfTSPPvsswP6HFdccUXWrFnzlutf/epXs2TJkgF9LoBfJn5PNcAg8Kd/+qd5+umn82d/9mcZO3ZskuSJJ57oj+D3v//9dV4IwM8iqgHqbPfu3fmbv/mb/PM//3Oampr6r0+ePDnz5s3La6+9liR5/vnns2TJkuzbty+VSiWXXXZZZs6cmSS5//77c++992bYsGF53/vel0WLFuWEE05IR0dH5s2bl127duX9739/9uzZc8j7nnnmmdx+++157bXXMmLEiHzpS19Kc3NzXn311XzlK1/J9u3bs2/fvhx55JFZtmxZTjzxxHz605/OpEmT8u1vfzsvvvhiJk+enJtuuinDhvkfpMAvJlENUGff+c53ctJJJ70pqH/qp9F84MCBXHnllbn++uvT0tKSjo6OXHDBBRk/fny6u7vzta99Lffff3+OPfbYrFmzJldddVW++c1vZsmSJTn55JPzpS99Kdu3b+//fgdr7969ufrqq/OXf/mXOfnkk/P888/n4osvzqpVq/Ld7343Rx11VO6///4kyeLFi3Pfffdl0aJFSZL/+q//yr333ptXX30155xzTp566qmcfvrphX9bAIOTqAaos//5wbZdXV2ZPXt2kvQH6bnnnpuenp60tLQkSUaPHp2Wlpb867/+a7q7uzN9+vQce+yxSZK2trbccsst2bFjRzZt2pQ/+qM/SpKMHz8+p5122iFt27JlS37lV34lJ598cpLkgx/8YH7jN34jTz31VGbNmpVx48bl3nvvzfbt2/PUU0/llFNO6f/aj3/84xk2bFhGjRqV8ePHZ//+/T/fXxDAECCqAersIx/5SF544YXs3bs3xxxzTEaNGpV169Yl+ckPEO7duze9vb2pVCpv+rpqtZoDBw6kr6/vLd/zp/dVKpU3RXtDw6G97P+s5/3617+elStXZvbs2Wltbc3RRx+dHTt29D/u8MMP7//z/9wB8IvG4TaAOhs9enTmzJmTL37xi/nRj37Uf33nzp359re/nWHDhuXEE09MQ0NDHnrooSRJR0dHHnzwwZxxxhk566yzsnHjxrz00ktJkpdNwCYAAAE+SURBVNWrV+foo4/O+PHjc9ZZZ/Ufz/jRj36Uf/u3fzukbZMmTcoPfvCDbNmyJclPznU//fTTOfXUU/PYY4/lvPPOywUXXJATTjghjzzySHp7ewfirwRgyPFONcAgcM0112T9+vW59tpr89prr+WVV17Je9/73kyfPj2zZ8/OiBEjsmLFitx888356le/mt7e3lx11VX9Z5QvueSSfOYzn0lfX1+OPfbY3HXXXRk2bFi+/OUv54Ybbsg555yTMWPGZMKECe+4YeXKlVm7dm3/7Q9/+MP5xje+kTvuuCM33XRTuru7U6lUcuutt+aEE07IZZddlsWLF2fVqlVJfhLg3/ve92r7FwUwSFWq/n8cAAAUcfwDAAAKiWoAACgkqgEAoJCoBgCAQqIaAAAKiWoAACgkqgEAoJCoBgCAQv8XDAVA+9nVTCsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#check if the dataset is balanced or not\n",
    "\n",
    "sns.countplot(df['Good Loan'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "id": "xIA5J087LVop"
   },
   "outputs": [],
   "source": [
    "#Balanced\n",
    "\n",
    "## next check the relation of all cat features wrt to dependent variable\n",
    "\n",
    "features = df.select_dtypes(exclude='number').drop('Good Loan', 1).columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "id": "Qx6EHmKZLYfX"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for feat in features:\n",
    "    sns.countplot(df[feat], hue = df['Good Loan'])\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "id": "6MR-OMOiLZ4X"
   },
   "outputs": [],
   "source": [
    "# since it is difficult to figure out which all features are important, we will go ahead with feature selection\n",
    "\n",
    "# but before that we need to encode\n",
    "\n",
    "df = pd.get_dummies(df,drop_first= True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "id": "Un3S_x0sLgE8"
   },
   "outputs": [],
   "source": [
    "df = df.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "MWuqCQitLjiS",
    "outputId": "04d36c0a-5ae9-4185-9317-fef463aa64ca"
   },
   "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>Loan Amount</th>\n",
       "      <th>Age</th>\n",
       "      <th>Type of Account_Type A</th>\n",
       "      <th>Type of Account_Type B</th>\n",
       "      <th>Type of Account_Type C</th>\n",
       "      <th>Account History_Good</th>\n",
       "      <th>Account History_critical</th>\n",
       "      <th>Reason for the Loan_All other</th>\n",
       "      <th>Reason for the Loan_Buying a New Car</th>\n",
       "      <th>Reason for the Loan_Buying a Used Car</th>\n",
       "      <th>...</th>\n",
       "      <th>Security / Collateral_Property - Real Estate</th>\n",
       "      <th>Security / Collateral_Savings Account</th>\n",
       "      <th>Security / Collateral_Vehicle</th>\n",
       "      <th>Residence Status_House Owner</th>\n",
       "      <th>Residence Status_Renting</th>\n",
       "      <th>Job_Professional / Management</th>\n",
       "      <th>Job_Services</th>\n",
       "      <th>Job_Skilled Labor</th>\n",
       "      <th>Completed Other loan?_Yes</th>\n",
       "      <th>Good Loan_Yes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2116.0</td>\n",
       "      <td>47.0</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",
       "      <td>0</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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2438.0</td>\n",
       "      <td>52.0</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",
       "      <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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2144.0</td>\n",
       "      <td>48.0</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",
       "      <td>1</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>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2073.0</td>\n",
       "      <td>25.0</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",
       "      <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>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1276.0</td>\n",
       "      <td>59.0</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",
       "      <td>0</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>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 38 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Loan Amount   Age  Type of Account_Type A  Type of Account_Type B  \\\n",
       "0       2116.0  47.0                       0                       0   \n",
       "1       2438.0  52.0                       0                       0   \n",
       "2       2144.0  48.0                       0                       0   \n",
       "3       2073.0  25.0                       0                       0   \n",
       "4       1276.0  59.0                       0                       0   \n",
       "\n",
       "   Type of Account_Type C  Account History_Good  Account History_critical  \\\n",
       "0                       0                     0                         1   \n",
       "1                       0                     0                         1   \n",
       "2                       0                     0                         1   \n",
       "3                       0                     0                         1   \n",
       "4                       0                     0                         1   \n",
       "\n",
       "   Reason for the Loan_All other  Reason for the Loan_Buying a New Car  \\\n",
       "0                              0                                     0   \n",
       "1                              0                                     0   \n",
       "2                              0                                     1   \n",
       "3                              0                                     0   \n",
       "4                              0                                     0   \n",
       "\n",
       "   Reason for the Loan_Buying a Used Car  ...  \\\n",
       "0                                      0  ...   \n",
       "1                                      0  ...   \n",
       "2                                      0  ...   \n",
       "3                                      0  ...   \n",
       "4                                      0  ...   \n",
       "\n",
       "   Security / Collateral_Property - Real Estate  \\\n",
       "0                                             1   \n",
       "1                                             0   \n",
       "2                                             0   \n",
       "3                                             0   \n",
       "4                                             0   \n",
       "\n",
       "   Security / Collateral_Savings Account  Security / Collateral_Vehicle  \\\n",
       "0                                      0                              0   \n",
       "1                                      1                              0   \n",
       "2                                      0                              1   \n",
       "3                                      0                              1   \n",
       "4                                      0                              0   \n",
       "\n",
       "   Residence Status_House Owner  Residence Status_Renting  \\\n",
       "0                             1                         0   \n",
       "1                             1                         0   \n",
       "2                             1                         0   \n",
       "3                             1                         0   \n",
       "4                             0                         1   \n",
       "\n",
       "   Job_Professional / Management  Job_Services  Job_Skilled Labor  \\\n",
       "0                              0             1                  0   \n",
       "1                              0             0                  1   \n",
       "2                              0             0                  1   \n",
       "3                              0             0                  1   \n",
       "4                              0             1                  0   \n",
       "\n",
       "   Completed Other loan?_Yes  Good Loan_Yes  \n",
       "0                          0              0  \n",
       "1                          0              0  \n",
       "2                          0              0  \n",
       "3                          1              0  \n",
       "4                          0              0  \n",
       "\n",
       "[5 rows x 38 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "R60Mr3D1LlGq",
    "outputId": "523b9036-ec2b-486f-cddc-2546fea8e1f6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1000, 37)\n",
      "(1000,)\n"
     ]
    }
   ],
   "source": [
    "### Feature selection\n",
    "\n",
    "# Split the data into X & y\n",
    "\n",
    "X = df.drop('Good Loan_Yes', axis = 1)\n",
    "y = df['Good Loan_Yes']\n",
    "print(X.shape)\n",
    "print(y.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "1Doy5CF3L2KO",
    "outputId": "ab4dc392-61bb-48cb-cfe9-ccdc76454288"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier(criterion='entropy', max_depth=10, random_state=15)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Run a Tree-based estimators (i.e. decision trees & random forests)\n",
    "\n",
    "dt = DecisionTreeClassifier(random_state=15, criterion = 'entropy', max_depth = 10)\n",
    "dt.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "FSruRrt7L2t7",
    "outputId": "1ace68c9-02bb-459e-fe9f-3158ea999eea"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The feature importance for Loan Amount is : 0.11851091312453825\n",
      "The feature importance for Age is : 0.11927203129113106\n",
      "The feature importance for Type of Account_Type A is : 0.05246414016750868\n",
      "The feature importance for Type of Account_Type B is : 0.029425472198296755\n",
      "The feature importance for Type of Account_Type C is : 0.04028580809112482\n",
      "The feature importance for Account History_Good is : 0.1858542283713255\n",
      "The feature importance for Account History_critical is : 0.006182710969959951\n",
      "The feature importance for Reason for the Loan_All other is : 0.002740052759790043\n",
      "The feature importance for Reason for the Loan_Buying a New Car is : 0.00701690622691178\n",
      "The feature importance for Reason for the Loan_Buying a Used Car is : 0.010037513584240801\n",
      "The feature importance for Reason for the Loan_Home Devices is : 0.0\n",
      "The feature importance for Reason for the Loan_Home furniture is : 0.008468038448578364\n",
      "The feature importance for Reason for the Loan_Learning / Edu purposes is : 0.00350845311345589\n",
      "The feature importance for Reason for the Loan_Renovation is : 0.007162743178725006\n",
      "The feature importance for Reason for the Loan_Support for Business is : 0.028423217309499278\n",
      "The feature importance for Reason for the Loan_TV is : 0.01386654889404015\n",
      "The feature importance for Account Savings_1000+ is : 0.056432381437204006\n",
      "The feature importance for Account Savings_200-500 is : 0.06474525463854244\n",
      "The feature importance for Account Savings_500-1000 is : 0.05221325572734246\n",
      "The feature importance for Account Savings_No Data is : 0.04876653976573717\n",
      "The feature importance for Employment History_2-5 Years is : 0.004895831035197403\n",
      "The feature importance for Employment History_5-7 Years is : 0.0030423380686068666\n",
      "The feature importance for Employment History_7+ Years is : 0.005994577109130236\n",
      "The feature importance for Employment History_Unemployed is : 0.009214629177417043\n",
      "The feature importance for Individual Stauts_Female is : 0.0\n",
      "The feature importance for Individual Stauts_Male is : 0.002576478326388624\n",
      "The feature importance for Individual Stauts_Married is : 0.007199814449753761\n",
      "The feature importance for Other Loans_Yes is : 0.009525627750780107\n",
      "The feature importance for Security / Collateral_Property - Real Estate is : 0.021371990485502996\n",
      "The feature importance for Security / Collateral_Savings Account is : 0.0032907936243295206\n",
      "The feature importance for Security / Collateral_Vehicle is : 0.024766888640272752\n",
      "The feature importance for Residence Status_House Owner is : 0.023229827433907755\n",
      "The feature importance for Residence Status_Renting is : 0.0\n",
      "The feature importance for Job_Professional / Management is : 0.004656646839861076\n",
      "The feature importance for Job_Services is : 0.007587936357664882\n",
      "The feature importance for Job_Skilled Labor is : 0.01314622730518346\n",
      "The feature importance for Completed Other loan?_Yes is : 0.004124184098051107\n"
     ]
    }
   ],
   "source": [
    "# Running Feature Importance\n",
    "\n",
    "fi_col = []\n",
    "fi = []\n",
    "\n",
    "for i,column in enumerate(df.drop('Good Loan_Yes', axis = 1)):\n",
    "    print('The feature importance for {} is : {}'.format(column, dt.feature_importances_[i]))\n",
    "    \n",
    "    fi_col.append(column)\n",
    "    fi.append(dt.feature_importances_[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "szXaG9l-L_hD",
    "outputId": "c1af9a34-cc7d-43cf-ade7-72acdd1c237c"
   },
   "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>index</th>\n",
       "      <th>Feature</th>\n",
       "      <th>Featrue_Imp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5</td>\n",
       "      <td>Account History_Good</td>\n",
       "      <td>0.185854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>Age</td>\n",
       "      <td>0.119272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>Loan Amount</td>\n",
       "      <td>0.118511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>17</td>\n",
       "      <td>Account Savings_200-500</td>\n",
       "      <td>0.064745</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>16</td>\n",
       "      <td>Account Savings_1000+</td>\n",
       "      <td>0.056432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2</td>\n",
       "      <td>Type of Account_Type A</td>\n",
       "      <td>0.052464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>18</td>\n",
       "      <td>Account Savings_500-1000</td>\n",
       "      <td>0.052213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>19</td>\n",
       "      <td>Account Savings_No Data</td>\n",
       "      <td>0.048767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>4</td>\n",
       "      <td>Type of Account_Type C</td>\n",
       "      <td>0.040286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>3</td>\n",
       "      <td>Type of Account_Type B</td>\n",
       "      <td>0.029425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>14</td>\n",
       "      <td>Reason for the Loan_Support for Business</td>\n",
       "      <td>0.028423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>30</td>\n",
       "      <td>Security / Collateral_Vehicle</td>\n",
       "      <td>0.024767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>31</td>\n",
       "      <td>Residence Status_House Owner</td>\n",
       "      <td>0.023230</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>28</td>\n",
       "      <td>Security / Collateral_Property - Real Estate</td>\n",
       "      <td>0.021372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>15</td>\n",
       "      <td>Reason for the Loan_TV</td>\n",
       "      <td>0.013867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>35</td>\n",
       "      <td>Job_Skilled Labor</td>\n",
       "      <td>0.013146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>9</td>\n",
       "      <td>Reason for the Loan_Buying a Used Car</td>\n",
       "      <td>0.010038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>27</td>\n",
       "      <td>Other Loans_Yes</td>\n",
       "      <td>0.009526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>23</td>\n",
       "      <td>Employment History_Unemployed</td>\n",
       "      <td>0.009215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>11</td>\n",
       "      <td>Reason for the Loan_Home furniture</td>\n",
       "      <td>0.008468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>34</td>\n",
       "      <td>Job_Services</td>\n",
       "      <td>0.007588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>26</td>\n",
       "      <td>Individual Stauts_Married</td>\n",
       "      <td>0.007200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>13</td>\n",
       "      <td>Reason for the Loan_Renovation</td>\n",
       "      <td>0.007163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>8</td>\n",
       "      <td>Reason for the Loan_Buying a New Car</td>\n",
       "      <td>0.007017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>6</td>\n",
       "      <td>Account History_critical</td>\n",
       "      <td>0.006183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>22</td>\n",
       "      <td>Employment History_7+ Years</td>\n",
       "      <td>0.005995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>20</td>\n",
       "      <td>Employment History_2-5 Years</td>\n",
       "      <td>0.004896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>33</td>\n",
       "      <td>Job_Professional / Management</td>\n",
       "      <td>0.004657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>36</td>\n",
       "      <td>Completed Other loan?_Yes</td>\n",
       "      <td>0.004124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>12</td>\n",
       "      <td>Reason for the Loan_Learning / Edu purposes</td>\n",
       "      <td>0.003508</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>29</td>\n",
       "      <td>Security / Collateral_Savings Account</td>\n",
       "      <td>0.003291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>21</td>\n",
       "      <td>Employment History_5-7 Years</td>\n",
       "      <td>0.003042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>7</td>\n",
       "      <td>Reason for the Loan_All other</td>\n",
       "      <td>0.002740</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>25</td>\n",
       "      <td>Individual Stauts_Male</td>\n",
       "      <td>0.002576</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    index                                       Feature  Featrue_Imp\n",
       "0       5                          Account History_Good     0.185854\n",
       "1       1                                           Age     0.119272\n",
       "2       0                                   Loan Amount     0.118511\n",
       "3      17                       Account Savings_200-500     0.064745\n",
       "4      16                         Account Savings_1000+     0.056432\n",
       "5       2                        Type of Account_Type A     0.052464\n",
       "6      18                      Account Savings_500-1000     0.052213\n",
       "7      19                       Account Savings_No Data     0.048767\n",
       "8       4                        Type of Account_Type C     0.040286\n",
       "9       3                        Type of Account_Type B     0.029425\n",
       "10     14      Reason for the Loan_Support for Business     0.028423\n",
       "11     30                 Security / Collateral_Vehicle     0.024767\n",
       "12     31                  Residence Status_House Owner     0.023230\n",
       "13     28  Security / Collateral_Property - Real Estate     0.021372\n",
       "14     15                        Reason for the Loan_TV     0.013867\n",
       "15     35                             Job_Skilled Labor     0.013146\n",
       "16      9         Reason for the Loan_Buying a Used Car     0.010038\n",
       "17     27                               Other Loans_Yes     0.009526\n",
       "18     23                 Employment History_Unemployed     0.009215\n",
       "19     11            Reason for the Loan_Home furniture     0.008468\n",
       "20     34                                  Job_Services     0.007588\n",
       "21     26                     Individual Stauts_Married     0.007200\n",
       "22     13                Reason for the Loan_Renovation     0.007163\n",
       "23      8          Reason for the Loan_Buying a New Car     0.007017\n",
       "24      6                      Account History_critical     0.006183\n",
       "25     22                   Employment History_7+ Years     0.005995\n",
       "26     20                  Employment History_2-5 Years     0.004896\n",
       "27     33                 Job_Professional / Management     0.004657\n",
       "28     36                     Completed Other loan?_Yes     0.004124\n",
       "29     12   Reason for the Loan_Learning / Edu purposes     0.003508\n",
       "30     29         Security / Collateral_Savings Account     0.003291\n",
       "31     21                  Employment History_5-7 Years     0.003042\n",
       "32      7                 Reason for the Loan_All other     0.002740\n",
       "33     25                        Individual Stauts_Male     0.002576"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Creating a dataframe\n",
    "\n",
    "imp_col = zip(fi_col, fi)\n",
    "imp_col = pd.DataFrame(imp_col, columns=['Feature','Featrue_Imp']).sort_values('Featrue_Imp', ascending = False).reset_index()\n",
    "imp_col = imp_col.iloc[:34,:]\n",
    "imp_col"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "id": "cqDYK_lKMC46"
   },
   "outputs": [],
   "source": [
    "X = df[imp_col['Feature'].values]\n",
    "y = df['Good Loan_Yes']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "id": "FP1EVB01YTp9"
   },
   "outputs": [],
   "source": [
    "from scipy.stats import zscore"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "id": "E3nuaD3gYgno"
   },
   "outputs": [],
   "source": [
    "X_new = X.copy()\n",
    "X_new[X.columns] = zscore(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 164
    },
    "id": "hC_fFYqRYuII",
    "outputId": "11ebc7c8-042f-4e18-9650-4da2cd8a0c45"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Account History_Good</th>\n",
       "      <th>Age</th>\n",
       "      <th>Loan Amount</th>\n",
       "      <th>Account Savings_200-500</th>\n",
       "      <th>Account Savings_1000+</th>\n",
       "      <th>Type of Account_Type A</th>\n",
       "      <th>Account Savings_500-1000</th>\n",
       "      <th>Account Savings_No Data</th>\n",
       "      <th>Type of Account_Type C</th>\n",
       "      <th>Type of Account_Type B</th>\n",
       "      <th>...</th>\n",
       "      <th>Account History_critical</th>\n",
       "      <th>Employment History_7+ Years</th>\n",
       "      <th>Employment History_2-5 Years</th>\n",
       "      <th>Job_Professional / Management</th>\n",
       "      <th>Completed Other loan?_Yes</th>\n",
       "      <th>Reason for the Loan_Learning / Edu purposes</th>\n",
       "      <th>Security / Collateral_Savings Account</th>\n",
       "      <th>Employment History_5-7 Years</th>\n",
       "      <th>Reason for the Loan_All other</th>\n",
       "      <th>Individual Stauts_Male</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.151339</td>\n",
       "      <td>1.004327</td>\n",
       "      <td>-0.413152</td>\n",
       "      <td>-0.338862</td>\n",
       "      <td>-0.224544</td>\n",
       "      <td>-0.524891</td>\n",
       "      <td>-0.259299</td>\n",
       "      <td>-0.473276</td>\n",
       "      <td>-0.687571</td>\n",
       "      <td>-0.67028</td>\n",
       "      <td>...</td>\n",
       "      <td>1.553374</td>\n",
       "      <td>-0.581969</td>\n",
       "      <td>-0.716142</td>\n",
       "      <td>-0.416784</td>\n",
       "      <td>-0.198749</td>\n",
       "      <td>4.358899</td>\n",
       "      <td>-0.549621</td>\n",
       "      <td>2.17879</td>\n",
       "      <td>-0.110208</td>\n",
       "      <td>0.908195</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 34 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Account History_Good       Age  Loan Amount  Account Savings_200-500  \\\n",
       "0             -1.151339  1.004327    -0.413152                -0.338862   \n",
       "\n",
       "   Account Savings_1000+  Type of Account_Type A  Account Savings_500-1000  \\\n",
       "0              -0.224544               -0.524891                 -0.259299   \n",
       "\n",
       "   Account Savings_No Data  Type of Account_Type C  Type of Account_Type B  \\\n",
       "0                -0.473276               -0.687571                -0.67028   \n",
       "\n",
       "   ...  Account History_critical  Employment History_7+ Years  \\\n",
       "0  ...                  1.553374                    -0.581969   \n",
       "\n",
       "   Employment History_2-5 Years  Job_Professional / Management  \\\n",
       "0                     -0.716142                      -0.416784   \n",
       "\n",
       "   Completed Other loan?_Yes  Reason for the Loan_Learning / Edu purposes  \\\n",
       "0                  -0.198749                                     4.358899   \n",
       "\n",
       "   Security / Collateral_Savings Account  Employment History_5-7 Years  \\\n",
       "0                              -0.549621                       2.17879   \n",
       "\n",
       "   Reason for the Loan_All other  Individual Stauts_Male  \n",
       "0                      -0.110208                0.908195  \n",
       "\n",
       "[1 rows x 34 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_new.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "xfLMMcOoMFxd",
    "outputId": "7c294561-fe18-4d05-90e9-a7be2f23e03b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(800, 34)\n",
      "(200, 34)\n",
      "(800,)\n",
      "(200,)\n"
     ]
    }
   ],
   "source": [
    "# Train test validation\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_new, y, train_size = 0.8, test_size=0.2, random_state=15)\n",
    "\n",
    "print(X_train.shape)\n",
    "print(X_test.shape)\n",
    "\n",
    "print(y_train.shape)\n",
    "print(y_test.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "id": "_p3wMlZ8MP8S"
   },
   "outputs": [],
   "source": [
    "## Building an ANN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "id": "_aDXFe7AMSv-"
   },
   "outputs": [],
   "source": [
    "import keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense,Dropout\n",
    "from keras.optimizers import Adam"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "id": "XKkfPKNgROOI"
   },
   "outputs": [],
   "source": [
    "classifier = Sequential()\n",
    "\n",
    "# First HL with 6 neurons and input to these is 34 features\n",
    "classifier.add(Dense(units=10, kernel_initializer='he_uniform',activation='relu',input_dim=34 ))\n",
    "classifier.add(Dropout(0.4))\n",
    "# Second HL\n",
    "classifier.add(Dense(units=10,kernel_initializer='he_uniform',activation='relu'))\n",
    "classifier.add(Dropout(0.4))\n",
    "# Third HL\n",
    "classifier.add(Dense(units=10,kernel_initializer='he_uniform',activation='relu'))\n",
    "classifier.add(Dropout(0.4))\n",
    "# Output Layer\n",
    "\n",
    "classifier.add(Dense(units=1,kernel_initializer='he_uniform',activation='sigmoid'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "id": "5ShHZ2fHSsRe"
   },
   "outputs": [],
   "source": [
    "\n",
    "# Compile the model\n",
    "\n",
    "classifier.compile(optimizer = 'adam',loss='binary_crossentropy',metrics=['accuracy'])\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "hRrLcMfNSunO",
    "outputId": "0ee63e96-d991-40a3-911b-124439293fe8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "54/54 [==============================] - 9s 91ms/step - loss: 1.5149 - accuracy: 0.4973 - val_loss: 0.8496 - val_accuracy: 0.4583\n",
      "Epoch 2/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 1.1369 - accuracy: 0.4749 - val_loss: 0.7564 - val_accuracy: 0.5189\n",
      "Epoch 3/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.9904 - accuracy: 0.4753 - val_loss: 0.7201 - val_accuracy: 0.5189\n",
      "Epoch 4/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.8151 - accuracy: 0.4969 - val_loss: 0.7030 - val_accuracy: 0.5152\n",
      "Epoch 5/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.8164 - accuracy: 0.4935 - val_loss: 0.6934 - val_accuracy: 0.5265\n",
      "Epoch 6/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.7868 - accuracy: 0.4719 - val_loss: 0.6887 - val_accuracy: 0.5455\n",
      "Epoch 7/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.7841 - accuracy: 0.4874 - val_loss: 0.6864 - val_accuracy: 0.5568\n",
      "Epoch 8/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.7391 - accuracy: 0.5400 - val_loss: 0.6837 - val_accuracy: 0.5530\n",
      "Epoch 9/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6893 - accuracy: 0.5996 - val_loss: 0.6823 - val_accuracy: 0.5417\n",
      "Epoch 10/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6942 - accuracy: 0.5439 - val_loss: 0.6809 - val_accuracy: 0.5417\n",
      "Epoch 11/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.7452 - accuracy: 0.5033 - val_loss: 0.6806 - val_accuracy: 0.5417\n",
      "Epoch 12/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.7358 - accuracy: 0.4721 - val_loss: 0.6808 - val_accuracy: 0.5568\n",
      "Epoch 13/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.7069 - accuracy: 0.5463 - val_loss: 0.6805 - val_accuracy: 0.5492\n",
      "Epoch 14/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6881 - accuracy: 0.5398 - val_loss: 0.6798 - val_accuracy: 0.5455\n",
      "Epoch 15/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6925 - accuracy: 0.5287 - val_loss: 0.6797 - val_accuracy: 0.5530\n",
      "Epoch 16/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.6519 - accuracy: 0.5994 - val_loss: 0.6794 - val_accuracy: 0.5455\n",
      "Epoch 17/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6828 - accuracy: 0.5305 - val_loss: 0.6791 - val_accuracy: 0.5417\n",
      "Epoch 18/100\n",
      "54/54 [==============================] - 0s 3ms/step - loss: 0.6767 - accuracy: 0.5713 - val_loss: 0.6785 - val_accuracy: 0.5530\n",
      "Epoch 19/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6770 - accuracy: 0.5523 - val_loss: 0.6780 - val_accuracy: 0.5530\n",
      "Epoch 20/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6822 - accuracy: 0.5631 - val_loss: 0.6772 - val_accuracy: 0.5492\n",
      "Epoch 21/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6676 - accuracy: 0.5530 - val_loss: 0.6760 - val_accuracy: 0.5492\n",
      "Epoch 22/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6905 - accuracy: 0.5343 - val_loss: 0.6762 - val_accuracy: 0.5568\n",
      "Epoch 23/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6776 - accuracy: 0.5422 - val_loss: 0.6753 - val_accuracy: 0.5682\n",
      "Epoch 24/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6942 - accuracy: 0.5564 - val_loss: 0.6750 - val_accuracy: 0.5606\n",
      "Epoch 25/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6792 - accuracy: 0.5575 - val_loss: 0.6733 - val_accuracy: 0.5720\n",
      "Epoch 26/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6694 - accuracy: 0.5808 - val_loss: 0.6716 - val_accuracy: 0.5720\n",
      "Epoch 27/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6730 - accuracy: 0.5746 - val_loss: 0.6698 - val_accuracy: 0.5758\n",
      "Epoch 28/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6621 - accuracy: 0.5691 - val_loss: 0.6668 - val_accuracy: 0.5758\n",
      "Epoch 29/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6632 - accuracy: 0.5909 - val_loss: 0.6654 - val_accuracy: 0.5795\n",
      "Epoch 30/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6663 - accuracy: 0.5505 - val_loss: 0.6644 - val_accuracy: 0.5795\n",
      "Epoch 31/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6607 - accuracy: 0.5703 - val_loss: 0.6612 - val_accuracy: 0.5833\n",
      "Epoch 32/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6649 - accuracy: 0.5706 - val_loss: 0.6584 - val_accuracy: 0.5909\n",
      "Epoch 33/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6617 - accuracy: 0.5858 - val_loss: 0.6538 - val_accuracy: 0.5947\n",
      "Epoch 34/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6754 - accuracy: 0.5283 - val_loss: 0.6517 - val_accuracy: 0.5871\n",
      "Epoch 35/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6506 - accuracy: 0.5603 - val_loss: 0.6487 - val_accuracy: 0.5871\n",
      "Epoch 36/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6784 - accuracy: 0.5611 - val_loss: 0.6451 - val_accuracy: 0.5795\n",
      "Epoch 37/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6505 - accuracy: 0.6093 - val_loss: 0.6376 - val_accuracy: 0.5985\n",
      "Epoch 38/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6301 - accuracy: 0.6214 - val_loss: 0.6316 - val_accuracy: 0.6250\n",
      "Epoch 39/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6459 - accuracy: 0.5750 - val_loss: 0.6266 - val_accuracy: 0.6288\n",
      "Epoch 40/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.6318 - accuracy: 0.6146 - val_loss: 0.6227 - val_accuracy: 0.6439\n",
      "Epoch 41/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6349 - accuracy: 0.5939 - val_loss: 0.6171 - val_accuracy: 0.6591\n",
      "Epoch 42/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6157 - accuracy: 0.6437 - val_loss: 0.6129 - val_accuracy: 0.6667\n",
      "Epoch 43/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6543 - accuracy: 0.5998 - val_loss: 0.6106 - val_accuracy: 0.6705\n",
      "Epoch 44/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5962 - accuracy: 0.6186 - val_loss: 0.6053 - val_accuracy: 0.6780\n",
      "Epoch 45/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5784 - accuracy: 0.6794 - val_loss: 0.5995 - val_accuracy: 0.6856\n",
      "Epoch 46/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6535 - accuracy: 0.6086 - val_loss: 0.5976 - val_accuracy: 0.7008\n",
      "Epoch 47/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.6062 - accuracy: 0.6128 - val_loss: 0.5939 - val_accuracy: 0.6894\n",
      "Epoch 48/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6290 - accuracy: 0.6412 - val_loss: 0.5890 - val_accuracy: 0.6856\n",
      "Epoch 49/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.6089 - accuracy: 0.6103 - val_loss: 0.5840 - val_accuracy: 0.6932\n",
      "Epoch 50/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5798 - accuracy: 0.6273 - val_loss: 0.5796 - val_accuracy: 0.6970\n",
      "Epoch 51/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.6010 - accuracy: 0.6458 - val_loss: 0.5772 - val_accuracy: 0.6894\n",
      "Epoch 52/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.5827 - accuracy: 0.6654 - val_loss: 0.5675 - val_accuracy: 0.6818\n",
      "Epoch 53/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5952 - accuracy: 0.6911 - val_loss: 0.5660 - val_accuracy: 0.7045\n",
      "Epoch 54/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5671 - accuracy: 0.6557 - val_loss: 0.5650 - val_accuracy: 0.7045\n",
      "Epoch 55/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5996 - accuracy: 0.6326 - val_loss: 0.5620 - val_accuracy: 0.7008\n",
      "Epoch 56/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5468 - accuracy: 0.7307 - val_loss: 0.5578 - val_accuracy: 0.7159\n",
      "Epoch 57/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5658 - accuracy: 0.6787 - val_loss: 0.5556 - val_accuracy: 0.7159\n",
      "Epoch 58/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5486 - accuracy: 0.7150 - val_loss: 0.5507 - val_accuracy: 0.7008\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 59/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5269 - accuracy: 0.7041 - val_loss: 0.5452 - val_accuracy: 0.7045\n",
      "Epoch 60/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5281 - accuracy: 0.7250 - val_loss: 0.5430 - val_accuracy: 0.7045\n",
      "Epoch 61/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5439 - accuracy: 0.7466 - val_loss: 0.5401 - val_accuracy: 0.6970\n",
      "Epoch 62/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5762 - accuracy: 0.6809 - val_loss: 0.5404 - val_accuracy: 0.7159\n",
      "Epoch 63/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5312 - accuracy: 0.7066 - val_loss: 0.5429 - val_accuracy: 0.7197\n",
      "Epoch 64/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5846 - accuracy: 0.6976 - val_loss: 0.5410 - val_accuracy: 0.7273\n",
      "Epoch 65/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5696 - accuracy: 0.6895 - val_loss: 0.5411 - val_accuracy: 0.7159\n",
      "Epoch 66/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5236 - accuracy: 0.7197 - val_loss: 0.5381 - val_accuracy: 0.7121\n",
      "Epoch 67/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5701 - accuracy: 0.7105 - val_loss: 0.5415 - val_accuracy: 0.7159\n",
      "Epoch 68/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.5493 - accuracy: 0.6994 - val_loss: 0.5405 - val_accuracy: 0.7121\n",
      "Epoch 69/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5628 - accuracy: 0.7192 - val_loss: 0.5382 - val_accuracy: 0.7121\n",
      "Epoch 70/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5080 - accuracy: 0.7297 - val_loss: 0.5317 - val_accuracy: 0.7159\n",
      "Epoch 71/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5032 - accuracy: 0.7355 - val_loss: 0.5299 - val_accuracy: 0.7121\n",
      "Epoch 72/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5160 - accuracy: 0.7242 - val_loss: 0.5291 - val_accuracy: 0.7121\n",
      "Epoch 73/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5330 - accuracy: 0.7086 - val_loss: 0.5282 - val_accuracy: 0.7197\n",
      "Epoch 74/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5335 - accuracy: 0.7467 - val_loss: 0.5319 - val_accuracy: 0.7121\n",
      "Epoch 75/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.5315 - accuracy: 0.7160 - val_loss: 0.5308 - val_accuracy: 0.7083\n",
      "Epoch 76/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5477 - accuracy: 0.6990 - val_loss: 0.5284 - val_accuracy: 0.7197\n",
      "Epoch 77/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5234 - accuracy: 0.7474 - val_loss: 0.5263 - val_accuracy: 0.7235\n",
      "Epoch 78/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5486 - accuracy: 0.6965 - val_loss: 0.5253 - val_accuracy: 0.7311\n",
      "Epoch 79/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5393 - accuracy: 0.7188 - val_loss: 0.5269 - val_accuracy: 0.7235\n",
      "Epoch 80/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.5278 - accuracy: 0.7162 - val_loss: 0.5245 - val_accuracy: 0.7159\n",
      "Epoch 81/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.5388 - accuracy: 0.7324 - val_loss: 0.5236 - val_accuracy: 0.7045\n",
      "Epoch 82/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.4983 - accuracy: 0.7921 - val_loss: 0.5200 - val_accuracy: 0.7197\n",
      "Epoch 83/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.4971 - accuracy: 0.7618 - val_loss: 0.5211 - val_accuracy: 0.7311\n",
      "Epoch 84/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5098 - accuracy: 0.7398 - val_loss: 0.5196 - val_accuracy: 0.7348\n",
      "Epoch 85/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5149 - accuracy: 0.7552 - val_loss: 0.5185 - val_accuracy: 0.7197\n",
      "Epoch 86/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5274 - accuracy: 0.7070 - val_loss: 0.5173 - val_accuracy: 0.7121\n",
      "Epoch 87/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5556 - accuracy: 0.6930 - val_loss: 0.5162 - val_accuracy: 0.7235\n",
      "Epoch 88/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5390 - accuracy: 0.7159 - val_loss: 0.5212 - val_accuracy: 0.7121\n",
      "Epoch 89/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.4967 - accuracy: 0.7566 - val_loss: 0.5226 - val_accuracy: 0.7045\n",
      "Epoch 90/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.4833 - accuracy: 0.7482 - val_loss: 0.5204 - val_accuracy: 0.7008\n",
      "Epoch 91/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.4896 - accuracy: 0.7601 - val_loss: 0.5214 - val_accuracy: 0.7008\n",
      "Epoch 92/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.4705 - accuracy: 0.7480 - val_loss: 0.5228 - val_accuracy: 0.6932\n",
      "Epoch 93/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.4708 - accuracy: 0.7326 - val_loss: 0.5227 - val_accuracy: 0.6932\n",
      "Epoch 94/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.5107 - accuracy: 0.7444 - val_loss: 0.5255 - val_accuracy: 0.6970\n",
      "Epoch 95/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.4759 - accuracy: 0.7950 - val_loss: 0.5265 - val_accuracy: 0.6818\n",
      "Epoch 96/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.4916 - accuracy: 0.7706 - val_loss: 0.5253 - val_accuracy: 0.6780\n",
      "Epoch 97/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.4525 - accuracy: 0.7439 - val_loss: 0.5258 - val_accuracy: 0.6818\n",
      "Epoch 98/100\n",
      "54/54 [==============================] - 0s 5ms/step - loss: 0.5174 - accuracy: 0.7206 - val_loss: 0.5250 - val_accuracy: 0.6894\n",
      "Epoch 99/100\n",
      "54/54 [==============================] - 0s 6ms/step - loss: 0.5157 - accuracy: 0.7511 - val_loss: 0.5238 - val_accuracy: 0.6856\n",
      "Epoch 100/100\n",
      "54/54 [==============================] - 0s 4ms/step - loss: 0.4767 - accuracy: 0.7772 - val_loss: 0.5226 - val_accuracy: 0.6970\n"
     ]
    }
   ],
   "source": [
    "## Fit the model\n",
    "model = classifier.fit(X_train,y_train,validation_split=0.33,batch_size=10,epochs=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "id": "HyMfA22DcUIR"
   },
   "outputs": [],
   "source": [
    "y_pred = classifier.predict(X_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "id": "Jubs8QrqdGCY"
   },
   "outputs": [],
   "source": [
    "y_pred = [1 if i >0.5 else 0 for i in y_pred]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Bik1XZWIcnLS",
    "outputId": "46ff99b1-b25c-4c6b-bbb6-f457371fb80e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.755"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "accuracy_score(y_pred,y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hyperparamter tuning "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "id": "Ocp3H_Ifcn7M"
   },
   "outputs": [],
   "source": [
    "from kerastuner.tuners import RandomSearch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0001"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "1e-4\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "def model(hp):\n",
    "    classifier = keras.Sequential()\n",
    "    \n",
    "    for i in range(hp.Int('layers',2,20)):\n",
    "        units = hp.Int('units',min_value=10,max_value=200, step=32)\n",
    "        classifier.add(Dense(units=units, activation='relu'))\n",
    "        \n",
    "    classifier.add(Dense(units=1,activation='sigmoid'))\n",
    "    hp_lr = hp.Choice('learning_rate', values=[0.01, 0.001, 0.0001])\n",
    "    classifier.compile(optimizer = Adam(learning_rate=hp_lr),loss='binary_crossentropy',metrics=['accuracy'])\n",
    "    \n",
    "    return classifier\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "tuner = RandomSearch(model, objective='val_accuracy', max_trials = 5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Search space summary\n",
      "Default search space size: 3\n",
      "layers (Int)\n",
      "{'default': None, 'conditions': [], 'min_value': 2, 'max_value': 20, 'step': 1, 'sampling': None}\n",
      "units (Int)\n",
      "{'default': None, 'conditions': [], 'min_value': 10, 'max_value': 200, 'step': 32, 'sampling': None}\n",
      "learning_rate (Choice)\n",
      "{'default': 0.01, 'conditions': [], 'values': [0.01, 0.001, 0.0001], 'ordered': True}\n"
     ]
    }
   ],
   "source": [
    "tuner.search_space_summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Search: Running Trial #1\n",
      "\n",
      "Hyperparameter    |Value             |Best Value So Far \n",
      "layers            |16                |?                 \n",
      "units             |138               |?                 \n",
      "learning_rate     |0.01              |?                 \n",
      "\n",
      "Epoch 1/5\n",
      "25/25 [==============================] - 4s 47ms/step - loss: 0.7047 - accuracy: 0.4903 - val_loss: 0.6915 - val_accuracy: 0.5300\n",
      "Epoch 2/5\n",
      "25/25 [==============================] - 0s 11ms/step - loss: 0.6886 - accuracy: 0.5587 - val_loss: 0.6913 - val_accuracy: 0.5300\n",
      "Epoch 3/5\n",
      "25/25 [==============================] - 0s 11ms/step - loss: 0.6899 - accuracy: 0.5414 - val_loss: 0.6916 - val_accuracy: 0.5300\n",
      "Epoch 4/5\n",
      "25/25 [==============================] - 0s 12ms/step - loss: 0.6884 - accuracy: 0.5457 - val_loss: 0.6893 - val_accuracy: 0.5300\n",
      "Epoch 5/5\n",
      "25/25 [==============================] - 0s 14ms/step - loss: 0.7076 - accuracy: 0.5246 - val_loss: 0.6934 - val_accuracy: 0.4700\n",
      "\n",
      "Trial 1 Complete [00h 00m 07s]\n",
      "val_accuracy: 0.5299999713897705\n",
      "\n",
      "Best val_accuracy So Far: 0.5299999713897705\n",
      "Total elapsed time: 00h 00m 07s\n",
      "\n",
      "Search: Running Trial #2\n",
      "\n",
      "Hyperparameter    |Value             |Best Value So Far \n",
      "layers            |8                 |16                \n",
      "units             |74                |138               \n",
      "learning_rate     |0.0001            |0.01              \n",
      "\n",
      "Epoch 1/5\n",
      "25/25 [==============================] - 2s 27ms/step - loss: 0.6886 - accuracy: 0.5330 - val_loss: 0.6833 - val_accuracy: 0.5300\n",
      "Epoch 2/5\n",
      "25/25 [==============================] - 0s 7ms/step - loss: 0.6859 - accuracy: 0.5117 - val_loss: 0.6769 - val_accuracy: 0.5300\n",
      "Epoch 3/5\n",
      "25/25 [==============================] - 0s 9ms/step - loss: 0.6764 - accuracy: 0.5194 - val_loss: 0.6670 - val_accuracy: 0.5300\n",
      "Epoch 4/5\n",
      "25/25 [==============================] - 0s 7ms/step - loss: 0.6642 - accuracy: 0.5462 - val_loss: 0.6519 - val_accuracy: 0.6100\n",
      "Epoch 5/5\n",
      "25/25 [==============================] - 0s 6ms/step - loss: 0.6362 - accuracy: 0.6536 - val_loss: 0.6291 - val_accuracy: 0.6850\n",
      "\n",
      "Trial 2 Complete [00h 00m 03s]\n",
      "val_accuracy: 0.6850000023841858\n",
      "\n",
      "Best val_accuracy So Far: 0.6850000023841858\n",
      "Total elapsed time: 00h 00m 11s\n",
      "\n",
      "Search: Running Trial #3\n",
      "\n",
      "Hyperparameter    |Value             |Best Value So Far \n",
      "layers            |14                |8                 \n",
      "units             |42                |74                \n",
      "learning_rate     |0.001             |0.0001            \n",
      "\n",
      "Epoch 1/5\n",
      "25/25 [==============================] - 3s 31ms/step - loss: 0.6925 - accuracy: 0.5129 - val_loss: 0.6856 - val_accuracy: 0.5300\n",
      "Epoch 2/5\n",
      "25/25 [==============================] - 0s 8ms/step - loss: 0.6699 - accuracy: 0.5541 - val_loss: 0.5998 - val_accuracy: 0.7050\n",
      "Epoch 3/5\n",
      "25/25 [==============================] - 0s 7ms/step - loss: 0.5898 - accuracy: 0.7355 - val_loss: 0.5310 - val_accuracy: 0.7350\n",
      "Epoch 4/5\n",
      "25/25 [==============================] - 0s 8ms/step - loss: 0.4717 - accuracy: 0.8047 - val_loss: 0.5348 - val_accuracy: 0.7550\n",
      "Epoch 5/5\n",
      "25/25 [==============================] - 0s 7ms/step - loss: 0.4448 - accuracy: 0.8143 - val_loss: 0.5083 - val_accuracy: 0.7550\n",
      "\n",
      "Trial 3 Complete [00h 00m 05s]\n",
      "val_accuracy: 0.7549999952316284\n",
      "\n",
      "Best val_accuracy So Far: 0.7549999952316284\n",
      "Total elapsed time: 00h 00m 16s\n",
      "\n",
      "Search: Running Trial #4\n",
      "\n",
      "Hyperparameter    |Value             |Best Value So Far \n",
      "layers            |14                |14                \n",
      "units             |138               |42                \n",
      "learning_rate     |0.01              |0.001             \n",
      "\n",
      "Epoch 1/5\n",
      "25/25 [==============================] - 3s 45ms/step - loss: 0.7094 - accuracy: 0.5233 - val_loss: 0.6848 - val_accuracy: 0.6650\n",
      "Epoch 2/5\n",
      "25/25 [==============================] - 0s 13ms/step - loss: 0.6689 - accuracy: 0.5849 - val_loss: 0.5210 - val_accuracy: 0.7050\n",
      "Epoch 3/5\n",
      "25/25 [==============================] - 0s 16ms/step - loss: 0.6304 - accuracy: 0.6946 - val_loss: 0.5685 - val_accuracy: 0.7300\n",
      "Epoch 4/5\n",
      "25/25 [==============================] - 0s 15ms/step - loss: 0.6162 - accuracy: 0.7088 - val_loss: 0.6734 - val_accuracy: 0.5300\n",
      "Epoch 5/5\n",
      "25/25 [==============================] - 0s 12ms/step - loss: 0.5474 - accuracy: 0.7277 - val_loss: 0.5962 - val_accuracy: 0.7200\n",
      "\n",
      "Trial 4 Complete [00h 00m 05s]\n",
      "val_accuracy: 0.7300000190734863\n",
      "\n",
      "Best val_accuracy So Far: 0.7549999952316284\n",
      "Total elapsed time: 00h 00m 22s\n",
      "\n",
      "Search: Running Trial #5\n",
      "\n",
      "Hyperparameter    |Value             |Best Value So Far \n",
      "layers            |10                |14                \n",
      "units             |170               |42                \n",
      "learning_rate     |0.001             |0.001             \n",
      "\n",
      "Epoch 1/5\n",
      "25/25 [==============================] - 2s 30ms/step - loss: 0.6753 - accuracy: 0.5596 - val_loss: 0.5155 - val_accuracy: 0.7600\n",
      "Epoch 2/5\n",
      "25/25 [==============================] - 0s 10ms/step - loss: 0.5264 - accuracy: 0.7236 - val_loss: 0.4624 - val_accuracy: 0.7600\n",
      "Epoch 3/5\n",
      "25/25 [==============================] - 0s 11ms/step - loss: 0.3841 - accuracy: 0.8226 - val_loss: 0.4900 - val_accuracy: 0.7550\n",
      "Epoch 4/5\n",
      "25/25 [==============================] - 0s 15ms/step - loss: 0.3638 - accuracy: 0.8199 - val_loss: 0.4398 - val_accuracy: 0.7900\n",
      "Epoch 5/5\n",
      "25/25 [==============================] - 0s 10ms/step - loss: 0.3003 - accuracy: 0.8673 - val_loss: 0.5603 - val_accuracy: 0.7600\n",
      "\n",
      "Trial 5 Complete [00h 00m 04s]\n",
      "val_accuracy: 0.7900000214576721\n",
      "\n",
      "Best val_accuracy So Far: 0.7900000214576721\n",
      "Total elapsed time: 00h 00m 27s\n",
      "INFO:tensorflow:Oracle triggered exit\n"
     ]
    }
   ],
   "source": [
    "tuner.search(X_train,y_train,epochs=5,validation_data=(X_test,y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Results summary\n",
      "Results in .\\untitled_project\n",
      "Showing 10 best trials\n",
      "Objective(name='val_accuracy', direction='max')\n",
      "Trial summary\n",
      "Hyperparameters:\n",
      "layers: 10\n",
      "units: 170\n",
      "learning_rate: 0.001\n",
      "Score: 0.7900000214576721\n",
      "Trial summary\n",
      "Hyperparameters:\n",
      "layers: 14\n",
      "units: 42\n",
      "learning_rate: 0.001\n",
      "Score: 0.7549999952316284\n",
      "Trial summary\n",
      "Hyperparameters:\n",
      "layers: 14\n",
      "units: 138\n",
      "learning_rate: 0.01\n",
      "Score: 0.7300000190734863\n",
      "Trial summary\n",
      "Hyperparameters:\n",
      "layers: 8\n",
      "units: 74\n",
      "learning_rate: 0.0001\n",
      "Score: 0.6850000023841858\n",
      "Trial summary\n",
      "Hyperparameters:\n",
      "layers: 16\n",
      "units: 138\n",
      "learning_rate: 0.01\n",
      "Score: 0.5299999713897705\n"
     ]
    }
   ],
   "source": [
    "tuner.results_summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Fittind the ANN again with the tuned parameters. layers=10, units = 170,lr=0.001\n",
    "\n",
    "classifier = Sequential()\n",
    "\n",
    "\n",
    "classifier.add(Dense(units=170, kernel_initializer='he_uniform',activation='relu',input_dim=34 ))\n",
    "classifier.add(Dropout(0.45))\n",
    "classifier.add(Dense(units=170,kernel_initializer='he_uniform',activation='relu'))\n",
    "classifier.add(Dropout(0.35))\n",
    "classifier.add(Dense(units=170,kernel_initializer='he_uniform',activation='relu'))\n",
    "classifier.add(Dropout(0.4))\n",
    "classifier.add(Dense(units=170,kernel_initializer='he_uniform',activation='relu'))\n",
    "classifier.add(Dropout(0.4))\n",
    "classifier.add(Dense(units=170,kernel_initializer='he_uniform',activation='relu'))\n",
    "classifier.add(Dropout(0.35))\n",
    "classifier.add(Dense(units=170,kernel_initializer='he_uniform',activation='relu'))\n",
    "classifier.add(Dropout(0.4))\n",
    "classifier.add(Dense(units=170,kernel_initializer='he_uniform',activation='relu'))\n",
    "classifier.add(Dropout(0.35))\n",
    "classifier.add(Dense(units=170,kernel_initializer='he_uniform',activation='relu'))\n",
    "classifier.add(Dropout(0.4))\n",
    "classifier.add(Dense(units=170,kernel_initializer='he_uniform',activation='relu'))\n",
    "classifier.add(Dropout(0.4))\n",
    "# Output Layer\n",
    "classifier.add(Dense(units=1,kernel_initializer='he_uniform',activation='sigmoid'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "\n",
    "classifier.compile(optimizer = Adam(learning_rate=0.001),loss='binary_crossentropy',metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "54/54 [==============================] - 3s 16ms/step - loss: 3.9379 - accuracy: 0.4255 - val_loss: 0.6918 - val_accuracy: 0.5076\n",
      "Epoch 2/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 1.4960 - accuracy: 0.5586 - val_loss: 0.6859 - val_accuracy: 0.5455\n",
      "Epoch 3/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 1.3938 - accuracy: 0.4788 - val_loss: 0.6864 - val_accuracy: 0.5568\n",
      "Epoch 4/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.8662 - accuracy: 0.5246 - val_loss: 0.6929 - val_accuracy: 0.5038\n",
      "Epoch 5/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.8445 - accuracy: 0.5647 - val_loss: 0.6924 - val_accuracy: 0.5227\n",
      "Epoch 6/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.9102 - accuracy: 0.4475 - val_loss: 0.6890 - val_accuracy: 0.5568\n",
      "Epoch 7/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.8332 - accuracy: 0.4787 - val_loss: 0.6903 - val_accuracy: 0.5492\n",
      "Epoch 8/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7833 - accuracy: 0.4831 - val_loss: 0.6909 - val_accuracy: 0.5530\n",
      "Epoch 9/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7467 - accuracy: 0.5259 - val_loss: 0.6901 - val_accuracy: 0.5530\n",
      "Epoch 10/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7199 - accuracy: 0.5113 - val_loss: 0.6895 - val_accuracy: 0.5530\n",
      "Epoch 11/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7391 - accuracy: 0.5031 - val_loss: 0.6896 - val_accuracy: 0.5530\n",
      "Epoch 12/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7646 - accuracy: 0.4714 - val_loss: 0.6903 - val_accuracy: 0.5530\n",
      "Epoch 13/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7086 - accuracy: 0.5569 - val_loss: 0.6903 - val_accuracy: 0.5530\n",
      "Epoch 14/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.7237 - accuracy: 0.5082 - val_loss: 0.6891 - val_accuracy: 0.5530\n",
      "Epoch 15/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.7416 - accuracy: 0.5135 - val_loss: 0.6889 - val_accuracy: 0.5530\n",
      "Epoch 16/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7215 - accuracy: 0.5120 - val_loss: 0.6903 - val_accuracy: 0.5530\n",
      "Epoch 17/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.6968 - accuracy: 0.5737 - val_loss: 0.6896 - val_accuracy: 0.5530\n",
      "Epoch 18/100\n",
      "54/54 [==============================] - 1s 9ms/step - loss: 0.7173 - accuracy: 0.5392 - val_loss: 0.6893 - val_accuracy: 0.5530\n",
      "Epoch 19/100\n",
      "54/54 [==============================] - 1s 11ms/step - loss: 0.6881 - accuracy: 0.5635 - val_loss: 0.6895 - val_accuracy: 0.5530\n",
      "Epoch 20/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.6937 - accuracy: 0.5495 - val_loss: 0.6904 - val_accuracy: 0.5530\n",
      "Epoch 21/100\n",
      "54/54 [==============================] - 1s 11ms/step - loss: 0.7241 - accuracy: 0.4887 - val_loss: 0.6898 - val_accuracy: 0.5530\n",
      "Epoch 22/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7030 - accuracy: 0.5014 - val_loss: 0.6895 - val_accuracy: 0.5530\n",
      "Epoch 23/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.6925 - accuracy: 0.5109 - val_loss: 0.6903 - val_accuracy: 0.5530\n",
      "Epoch 24/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.7009 - accuracy: 0.5308 - val_loss: 0.6905 - val_accuracy: 0.5530\n",
      "Epoch 25/100\n",
      "54/54 [==============================] - 1s 12ms/step - loss: 0.7179 - accuracy: 0.4942 - val_loss: 0.6900 - val_accuracy: 0.5530\n",
      "Epoch 26/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.7055 - accuracy: 0.5207 - val_loss: 0.6902 - val_accuracy: 0.5530\n",
      "Epoch 27/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6964 - accuracy: 0.5102 - val_loss: 0.6900 - val_accuracy: 0.5530\n",
      "Epoch 28/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6995 - accuracy: 0.5217 - val_loss: 0.6899 - val_accuracy: 0.5530\n",
      "Epoch 29/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6976 - accuracy: 0.5349 - val_loss: 0.6893 - val_accuracy: 0.5530\n",
      "Epoch 30/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.7124 - accuracy: 0.4918 - val_loss: 0.6896 - val_accuracy: 0.5530\n",
      "Epoch 31/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.7048 - accuracy: 0.5006 - val_loss: 0.6892 - val_accuracy: 0.5530\n",
      "Epoch 32/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6887 - accuracy: 0.5545 - val_loss: 0.6894 - val_accuracy: 0.5530\n",
      "Epoch 33/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.7029 - accuracy: 0.5032 - val_loss: 0.6898 - val_accuracy: 0.5530\n",
      "Epoch 34/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6982 - accuracy: 0.4891 - val_loss: 0.6898 - val_accuracy: 0.5530\n",
      "Epoch 35/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6953 - accuracy: 0.5300 - val_loss: 0.6896 - val_accuracy: 0.5530\n",
      "Epoch 36/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.7166 - accuracy: 0.4641 - val_loss: 0.6889 - val_accuracy: 0.5530\n",
      "Epoch 37/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6979 - accuracy: 0.5495 - val_loss: 0.6894 - val_accuracy: 0.5530\n",
      "Epoch 38/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6901 - accuracy: 0.5693 - val_loss: 0.6895 - val_accuracy: 0.5530\n",
      "Epoch 39/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.7002 - accuracy: 0.5354 - val_loss: 0.6878 - val_accuracy: 0.5530\n",
      "Epoch 40/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6859 - accuracy: 0.5395 - val_loss: 0.6870 - val_accuracy: 0.5530\n",
      "Epoch 41/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.6766 - accuracy: 0.5641 - val_loss: 0.6873 - val_accuracy: 0.5530\n",
      "Epoch 42/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.6981 - accuracy: 0.5474 - val_loss: 0.6888 - val_accuracy: 0.5530\n",
      "Epoch 43/100\n",
      "54/54 [==============================] - 1s 9ms/step - loss: 0.6914 - accuracy: 0.5571 - val_loss: 0.6882 - val_accuracy: 0.5530\n",
      "Epoch 44/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.6994 - accuracy: 0.5513 - val_loss: 0.6886 - val_accuracy: 0.5530\n",
      "Epoch 45/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7130 - accuracy: 0.4575 - val_loss: 0.6888 - val_accuracy: 0.5530\n",
      "Epoch 46/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.6900 - accuracy: 0.5475 - val_loss: 0.6882 - val_accuracy: 0.5530\n",
      "Epoch 47/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6981 - accuracy: 0.5043 - val_loss: 0.6878 - val_accuracy: 0.5530\n",
      "Epoch 48/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7094 - accuracy: 0.5014 - val_loss: 0.6874 - val_accuracy: 0.5530\n",
      "Epoch 49/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.6930 - accuracy: 0.5516 - val_loss: 0.6870 - val_accuracy: 0.5530\n",
      "Epoch 50/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.6949 - accuracy: 0.5333 - val_loss: 0.6876 - val_accuracy: 0.5530\n",
      "Epoch 51/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.6979 - accuracy: 0.5382 - val_loss: 0.6868 - val_accuracy: 0.5530\n",
      "Epoch 52/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.7023 - accuracy: 0.5201 - val_loss: 0.6871 - val_accuracy: 0.5530\n",
      "Epoch 53/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6836 - accuracy: 0.5503 - val_loss: 0.6875 - val_accuracy: 0.5530\n",
      "Epoch 54/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6751 - accuracy: 0.5566 - val_loss: 0.6862 - val_accuracy: 0.5568\n",
      "Epoch 55/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.7121 - accuracy: 0.5187 - val_loss: 0.6866 - val_accuracy: 0.5530\n",
      "Epoch 56/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.7009 - accuracy: 0.5312 - val_loss: 0.6842 - val_accuracy: 0.5606\n",
      "Epoch 57/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6914 - accuracy: 0.6052 - val_loss: 0.6861 - val_accuracy: 0.5530\n",
      "Epoch 58/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.6689 - accuracy: 0.6136 - val_loss: 0.6800 - val_accuracy: 0.5530\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 59/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.6942 - accuracy: 0.5788 - val_loss: 0.6809 - val_accuracy: 0.5530\n",
      "Epoch 60/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.6736 - accuracy: 0.5739 - val_loss: 0.6721 - val_accuracy: 0.6061\n",
      "Epoch 61/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.6732 - accuracy: 0.6066 - val_loss: 0.6683 - val_accuracy: 0.5871\n",
      "Epoch 62/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.6695 - accuracy: 0.6354 - val_loss: 0.6381 - val_accuracy: 0.7159\n",
      "Epoch 63/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.6874 - accuracy: 0.5968 - val_loss: 0.6684 - val_accuracy: 0.5568\n",
      "Epoch 64/100\n",
      "54/54 [==============================] - 1s 11ms/step - loss: 0.6441 - accuracy: 0.6356 - val_loss: 0.6473 - val_accuracy: 0.5682\n",
      "Epoch 65/100\n",
      "54/54 [==============================] - 1s 12ms/step - loss: 0.6275 - accuracy: 0.6413 - val_loss: 0.6309 - val_accuracy: 0.6098\n",
      "Epoch 66/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.6557 - accuracy: 0.6458 - val_loss: 0.5894 - val_accuracy: 0.7045\n",
      "Epoch 67/100\n",
      "54/54 [==============================] - 1s 11ms/step - loss: 0.6023 - accuracy: 0.6811 - val_loss: 0.5566 - val_accuracy: 0.7121\n",
      "Epoch 68/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.5871 - accuracy: 0.6966 - val_loss: 0.5565 - val_accuracy: 0.7311\n",
      "Epoch 69/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.5369 - accuracy: 0.6948 - val_loss: 0.5835 - val_accuracy: 0.7159\n",
      "Epoch 70/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.5910 - accuracy: 0.7384 - val_loss: 0.5535 - val_accuracy: 0.7348\n",
      "Epoch 71/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.5555 - accuracy: 0.7381 - val_loss: 0.5817 - val_accuracy: 0.6705\n",
      "Epoch 72/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.5322 - accuracy: 0.7327 - val_loss: 0.5285 - val_accuracy: 0.7500\n",
      "Epoch 73/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.5993 - accuracy: 0.7325 - val_loss: 0.5475 - val_accuracy: 0.7386\n",
      "Epoch 74/100\n",
      "54/54 [==============================] - 0s 7ms/step - loss: 0.5070 - accuracy: 0.7405 - val_loss: 0.5664 - val_accuracy: 0.7273\n",
      "Epoch 75/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.4756 - accuracy: 0.7643 - val_loss: 0.5350 - val_accuracy: 0.7273\n",
      "Epoch 76/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.4855 - accuracy: 0.7257 - val_loss: 0.5508 - val_accuracy: 0.7386\n",
      "Epoch 77/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.5136 - accuracy: 0.7716 - val_loss: 0.5395 - val_accuracy: 0.7386\n",
      "Epoch 78/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.4912 - accuracy: 0.7759 - val_loss: 0.5504 - val_accuracy: 0.7462\n",
      "Epoch 79/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.4363 - accuracy: 0.8022 - val_loss: 0.5622 - val_accuracy: 0.7121\n",
      "Epoch 80/100\n",
      "54/54 [==============================] - 1s 14ms/step - loss: 0.4386 - accuracy: 0.7764 - val_loss: 0.5445 - val_accuracy: 0.7311\n",
      "Epoch 81/100\n",
      "54/54 [==============================] - 1s 15ms/step - loss: 0.4689 - accuracy: 0.7502 - val_loss: 0.5358 - val_accuracy: 0.7424\n",
      "Epoch 82/100\n",
      "54/54 [==============================] - 1s 15ms/step - loss: 0.4910 - accuracy: 0.7966 - val_loss: 0.6268 - val_accuracy: 0.7348\n",
      "Epoch 83/100\n",
      "54/54 [==============================] - 1s 13ms/step - loss: 0.4014 - accuracy: 0.8097 - val_loss: 0.5341 - val_accuracy: 0.7538\n",
      "Epoch 84/100\n",
      "54/54 [==============================] - 1s 14ms/step - loss: 0.4638 - accuracy: 0.8203 - val_loss: 0.5371 - val_accuracy: 0.7424\n",
      "Epoch 85/100\n",
      "54/54 [==============================] - 1s 15ms/step - loss: 0.4156 - accuracy: 0.8178 - val_loss: 0.5846 - val_accuracy: 0.7348\n",
      "Epoch 86/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.4253 - accuracy: 0.8439 - val_loss: 0.5305 - val_accuracy: 0.7500\n",
      "Epoch 87/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.3886 - accuracy: 0.7942 - val_loss: 0.5613 - val_accuracy: 0.7386\n",
      "Epoch 88/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.4232 - accuracy: 0.8268 - val_loss: 0.5826 - val_accuracy: 0.7348\n",
      "Epoch 89/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.4237 - accuracy: 0.7960 - val_loss: 0.5778 - val_accuracy: 0.7462\n",
      "Epoch 90/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.3178 - accuracy: 0.8687 - val_loss: 0.5754 - val_accuracy: 0.7197\n",
      "Epoch 91/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.3942 - accuracy: 0.8114 - val_loss: 0.5218 - val_accuracy: 0.7273\n",
      "Epoch 92/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.3789 - accuracy: 0.8242 - val_loss: 0.5269 - val_accuracy: 0.7235\n",
      "Epoch 93/100\n",
      "54/54 [==============================] - 0s 8ms/step - loss: 0.4242 - accuracy: 0.8060 - val_loss: 0.5666 - val_accuracy: 0.7386\n",
      "Epoch 94/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.3500 - accuracy: 0.8392 - val_loss: 0.5350 - val_accuracy: 0.7273\n",
      "Epoch 95/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.4329 - accuracy: 0.8040 - val_loss: 0.5491 - val_accuracy: 0.7235\n",
      "Epoch 96/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.3702 - accuracy: 0.8443 - val_loss: 0.5371 - val_accuracy: 0.7197\n",
      "Epoch 97/100\n",
      "54/54 [==============================] - 1s 10ms/step - loss: 0.3608 - accuracy: 0.8108 - val_loss: 0.5290 - val_accuracy: 0.7121\n",
      "Epoch 98/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.3436 - accuracy: 0.8695 - val_loss: 0.5393 - val_accuracy: 0.7386\n",
      "Epoch 99/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.3524 - accuracy: 0.8369 - val_loss: 0.5528 - val_accuracy: 0.7424\n",
      "Epoch 100/100\n",
      "54/54 [==============================] - 0s 9ms/step - loss: 0.4007 - accuracy: 0.8230 - val_loss: 0.6590 - val_accuracy: 0.7424\n"
     ]
    }
   ],
   "source": [
    "## Fit the model\n",
    "model = classifier.fit(X_train,y_train,validation_split=0.33,batch_size=10,epochs=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.76"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = classifier.predict(X_test)\n",
    "y_pred = [1 if i >0.5 else 0 for i in y_pred]\n",
    "accuracy_score(y_pred,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "LoanPrediction_ANN.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}