ClubLoan.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9578, 14)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_df = pd.read_csv(\"loan_data.csv\")\n",
    "raw_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>credit.policy</th>\n",
       "      <th>int.rate</th>\n",
       "      <th>installment</th>\n",
       "      <th>log.annual.inc</th>\n",
       "      <th>dti</th>\n",
       "      <th>fico</th>\n",
       "      <th>days.with.cr.line</th>\n",
       "      <th>revol.bal</th>\n",
       "      <th>revol.util</th>\n",
       "      <th>inq.last.6mths</th>\n",
       "      <th>delinq.2yrs</th>\n",
       "      <th>pub.rec</th>\n",
       "      <th>not.fully.paid</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9.578000e+03</td>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9578.000000</td>\n",
       "      <td>9578.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.804970</td>\n",
       "      <td>0.122640</td>\n",
       "      <td>319.089413</td>\n",
       "      <td>10.932117</td>\n",
       "      <td>12.606679</td>\n",
       "      <td>710.846314</td>\n",
       "      <td>4560.767197</td>\n",
       "      <td>1.691396e+04</td>\n",
       "      <td>46.799236</td>\n",
       "      <td>1.577469</td>\n",
       "      <td>0.163708</td>\n",
       "      <td>0.062122</td>\n",
       "      <td>0.160054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.396245</td>\n",
       "      <td>0.026847</td>\n",
       "      <td>207.071301</td>\n",
       "      <td>0.614813</td>\n",
       "      <td>6.883970</td>\n",
       "      <td>37.970537</td>\n",
       "      <td>2496.930377</td>\n",
       "      <td>3.375619e+04</td>\n",
       "      <td>29.014417</td>\n",
       "      <td>2.200245</td>\n",
       "      <td>0.546215</td>\n",
       "      <td>0.262126</td>\n",
       "      <td>0.366676</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.060000</td>\n",
       "      <td>15.670000</td>\n",
       "      <td>7.547502</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>612.000000</td>\n",
       "      <td>178.958333</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.103900</td>\n",
       "      <td>163.770000</td>\n",
       "      <td>10.558414</td>\n",
       "      <td>7.212500</td>\n",
       "      <td>682.000000</td>\n",
       "      <td>2820.000000</td>\n",
       "      <td>3.187000e+03</td>\n",
       "      <td>22.600000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.122100</td>\n",
       "      <td>268.950000</td>\n",
       "      <td>10.928884</td>\n",
       "      <td>12.665000</td>\n",
       "      <td>707.000000</td>\n",
       "      <td>4139.958333</td>\n",
       "      <td>8.596000e+03</td>\n",
       "      <td>46.300000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.140700</td>\n",
       "      <td>432.762500</td>\n",
       "      <td>11.291293</td>\n",
       "      <td>17.950000</td>\n",
       "      <td>737.000000</td>\n",
       "      <td>5730.000000</td>\n",
       "      <td>1.824950e+04</td>\n",
       "      <td>70.900000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.216400</td>\n",
       "      <td>940.140000</td>\n",
       "      <td>14.528354</td>\n",
       "      <td>29.960000</td>\n",
       "      <td>827.000000</td>\n",
       "      <td>17639.958330</td>\n",
       "      <td>1.207359e+06</td>\n",
       "      <td>119.000000</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       credit.policy     int.rate  installment  log.annual.inc          dti  \\\n",
       "count    9578.000000  9578.000000  9578.000000     9578.000000  9578.000000   \n",
       "mean        0.804970     0.122640   319.089413       10.932117    12.606679   \n",
       "std         0.396245     0.026847   207.071301        0.614813     6.883970   \n",
       "min         0.000000     0.060000    15.670000        7.547502     0.000000   \n",
       "25%         1.000000     0.103900   163.770000       10.558414     7.212500   \n",
       "50%         1.000000     0.122100   268.950000       10.928884    12.665000   \n",
       "75%         1.000000     0.140700   432.762500       11.291293    17.950000   \n",
       "max         1.000000     0.216400   940.140000       14.528354    29.960000   \n",
       "\n",
       "              fico  days.with.cr.line     revol.bal   revol.util  \\\n",
       "count  9578.000000        9578.000000  9.578000e+03  9578.000000   \n",
       "mean    710.846314        4560.767197  1.691396e+04    46.799236   \n",
       "std      37.970537        2496.930377  3.375619e+04    29.014417   \n",
       "min     612.000000         178.958333  0.000000e+00     0.000000   \n",
       "25%     682.000000        2820.000000  3.187000e+03    22.600000   \n",
       "50%     707.000000        4139.958333  8.596000e+03    46.300000   \n",
       "75%     737.000000        5730.000000  1.824950e+04    70.900000   \n",
       "max     827.000000       17639.958330  1.207359e+06   119.000000   \n",
       "\n",
       "       inq.last.6mths  delinq.2yrs      pub.rec  not.fully.paid  \n",
       "count     9578.000000  9578.000000  9578.000000     9578.000000  \n",
       "mean         1.577469     0.163708     0.062122        0.160054  \n",
       "std          2.200245     0.546215     0.262126        0.366676  \n",
       "min          0.000000     0.000000     0.000000        0.000000  \n",
       "25%          0.000000     0.000000     0.000000        0.000000  \n",
       "50%          1.000000     0.000000     0.000000        0.000000  \n",
       "75%          2.000000     0.000000     0.000000        0.000000  \n",
       "max         33.000000    13.000000     5.000000        1.000000  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Getting a Statstical summary of data\n",
    "raw_df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 9578 entries, 0 to 9577\n",
      "Data columns (total 14 columns):\n",
      "credit.policy        9578 non-null int64\n",
      "purpose              9578 non-null object\n",
      "int.rate             9578 non-null float64\n",
      "installment          9578 non-null float64\n",
      "log.annual.inc       9578 non-null float64\n",
      "dti                  9578 non-null float64\n",
      "fico                 9578 non-null int64\n",
      "days.with.cr.line    9578 non-null float64\n",
      "revol.bal            9578 non-null int64\n",
      "revol.util           9578 non-null float64\n",
      "inq.last.6mths       9578 non-null int64\n",
      "delinq.2yrs          9578 non-null int64\n",
      "pub.rec              9578 non-null int64\n",
      "not.fully.paid       9578 non-null int64\n",
      "dtypes: float64(6), int64(7), object(1)\n",
      "memory usage: 1.0+ MB\n"
     ]
    }
   ],
   "source": [
    "# Getting the datatype and checking for null values if any\n",
    "raw_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "debt_consolidation    3957\n",
       "all_other             2331\n",
       "credit_card           1262\n",
       "home_improvement       629\n",
       "small_business         619\n",
       "major_purchase         437\n",
       "educational            343\n",
       "Name: purpose, dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_df['purpose'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>purpose_all_other</th>\n",
       "      <th>purpose_credit_card</th>\n",
       "      <th>purpose_debt_consolidation</th>\n",
       "      <th>purpose_educational</th>\n",
       "      <th>purpose_home_improvement</th>\n",
       "      <th>purpose_major_purchase</th>\n",
       "      <th>purpose_small_business</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   purpose_all_other  purpose_credit_card  purpose_debt_consolidation  \\\n",
       "0                  0                    0                           1   \n",
       "1                  0                    1                           0   \n",
       "2                  0                    0                           1   \n",
       "3                  0                    0                           1   \n",
       "4                  0                    1                           0   \n",
       "\n",
       "   purpose_educational  purpose_home_improvement  purpose_major_purchase  \\\n",
       "0                    0                         0                       0   \n",
       "1                    0                         0                       0   \n",
       "2                    0                         0                       0   \n",
       "3                    0                         0                       0   \n",
       "4                    0                         0                       0   \n",
       "\n",
       "   purpose_small_business  \n",
       "0                       0  \n",
       "1                       0  \n",
       "2                       0  \n",
       "3                       0  \n",
       "4                       0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Encoding the categorical values to numerical\n",
    "purpose_encoded = pd.get_dummies(raw_df['purpose'])\n",
    "purpose_encoded.columns = ['purpose_all_other','purpose_credit_card','purpose_debt_consolidation',\n",
    "                           'purpose_educational','purpose_home_improvement','purpose_major_purchase',\n",
    "                          'purpose_small_business']\n",
    "purpose_encoded.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>credit.policy</th>\n",
       "      <th>int.rate</th>\n",
       "      <th>installment</th>\n",
       "      <th>log.annual.inc</th>\n",
       "      <th>dti</th>\n",
       "      <th>fico</th>\n",
       "      <th>days.with.cr.line</th>\n",
       "      <th>revol.bal</th>\n",
       "      <th>revol.util</th>\n",
       "      <th>inq.last.6mths</th>\n",
       "      <th>delinq.2yrs</th>\n",
       "      <th>pub.rec</th>\n",
       "      <th>not.fully.paid</th>\n",
       "      <th>purpose_all_other</th>\n",
       "      <th>purpose_credit_card</th>\n",
       "      <th>purpose_debt_consolidation</th>\n",
       "      <th>purpose_educational</th>\n",
       "      <th>purpose_home_improvement</th>\n",
       "      <th>purpose_major_purchase</th>\n",
       "      <th>purpose_small_business</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.1189</td>\n",
       "      <td>829.10</td>\n",
       "      <td>11.350407</td>\n",
       "      <td>19.48</td>\n",
       "      <td>737</td>\n",
       "      <td>5639.958333</td>\n",
       "      <td>28854</td>\n",
       "      <td>52.1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0.1071</td>\n",
       "      <td>228.22</td>\n",
       "      <td>11.082143</td>\n",
       "      <td>14.29</td>\n",
       "      <td>707</td>\n",
       "      <td>2760.000000</td>\n",
       "      <td>33623</td>\n",
       "      <td>76.7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0.1357</td>\n",
       "      <td>366.86</td>\n",
       "      <td>10.373491</td>\n",
       "      <td>11.63</td>\n",
       "      <td>682</td>\n",
       "      <td>4710.000000</td>\n",
       "      <td>3511</td>\n",
       "      <td>25.6</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0.1008</td>\n",
       "      <td>162.34</td>\n",
       "      <td>11.350407</td>\n",
       "      <td>8.10</td>\n",
       "      <td>712</td>\n",
       "      <td>2699.958333</td>\n",
       "      <td>33667</td>\n",
       "      <td>73.2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0.1426</td>\n",
       "      <td>102.92</td>\n",
       "      <td>11.299732</td>\n",
       "      <td>14.97</td>\n",
       "      <td>667</td>\n",
       "      <td>4066.000000</td>\n",
       "      <td>4740</td>\n",
       "      <td>39.5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   credit.policy  int.rate  installment  log.annual.inc    dti  fico  \\\n",
       "0              1    0.1189       829.10       11.350407  19.48   737   \n",
       "1              1    0.1071       228.22       11.082143  14.29   707   \n",
       "2              1    0.1357       366.86       10.373491  11.63   682   \n",
       "3              1    0.1008       162.34       11.350407   8.10   712   \n",
       "4              1    0.1426       102.92       11.299732  14.97   667   \n",
       "\n",
       "   days.with.cr.line  revol.bal  revol.util  inq.last.6mths  delinq.2yrs  \\\n",
       "0        5639.958333      28854        52.1               0            0   \n",
       "1        2760.000000      33623        76.7               0            0   \n",
       "2        4710.000000       3511        25.6               1            0   \n",
       "3        2699.958333      33667        73.2               1            0   \n",
       "4        4066.000000       4740        39.5               0            1   \n",
       "\n",
       "   pub.rec  not.fully.paid  purpose_all_other  purpose_credit_card  \\\n",
       "0        0               0                  0                    0   \n",
       "1        0               0                  0                    1   \n",
       "2        0               0                  0                    0   \n",
       "3        0               0                  0                    0   \n",
       "4        0               0                  0                    1   \n",
       "\n",
       "   purpose_debt_consolidation  purpose_educational  purpose_home_improvement  \\\n",
       "0                           1                    0                         0   \n",
       "1                           0                    0                         0   \n",
       "2                           1                    0                         0   \n",
       "3                           1                    0                         0   \n",
       "4                           0                    0                         0   \n",
       "\n",
       "   purpose_major_purchase  purpose_small_business  \n",
       "0                       0                       0  \n",
       "1                       0                       0  \n",
       "2                       0                       0  \n",
       "3                       0                       0  \n",
       "4                       0                       0  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# adding the encoded columns to the df and droping the purpose column\n",
    "raw_df = raw_df.join(purpose_encoded)\n",
    "raw_df.drop(['purpose'],axis=1,inplace=True)\n",
    "raw_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    7710\n",
       "0    1868\n",
       "Name: credit.policy, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_df['credit.policy'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "c = ['credit.policy','int.rate','installment','log.annual.inc','dti','fico','days.with.cr.line','revol.bal','revol.util','inq.last.6mths','delinq.2yrs','pub.rec','not.fully.paid']\n",
    "temp_df = raw_df.loc[:,c]\n",
    "corr = temp_df.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x1440 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "mask = np.zeros_like(corr)\n",
    "mask[np.triu_indices_from(mask)] = True\n",
    "with sns.axes_style(\"white\"):\n",
    "    f, ax = plt.subplots(figsize=(25,20))\n",
    "    ax = sns.heatmap(corr, mask=mask, vmax=.3, square=True,annot=True,linewidth=.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "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>credit.policy</th>\n",
       "      <th>int.rate</th>\n",
       "      <th>installment</th>\n",
       "      <th>log.annual.inc</th>\n",
       "      <th>dti</th>\n",
       "      <th>days.with.cr.line</th>\n",
       "      <th>revol.bal</th>\n",
       "      <th>revol.util</th>\n",
       "      <th>inq.last.6mths</th>\n",
       "      <th>delinq.2yrs</th>\n",
       "      <th>pub.rec</th>\n",
       "      <th>not.fully.paid</th>\n",
       "      <th>purpose_all_other</th>\n",
       "      <th>purpose_credit_card</th>\n",
       "      <th>purpose_debt_consolidation</th>\n",
       "      <th>purpose_educational</th>\n",
       "      <th>purpose_home_improvement</th>\n",
       "      <th>purpose_major_purchase</th>\n",
       "      <th>purpose_small_business</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.1189</td>\n",
       "      <td>829.10</td>\n",
       "      <td>11.350407</td>\n",
       "      <td>19.48</td>\n",
       "      <td>5639.958333</td>\n",
       "      <td>28854</td>\n",
       "      <td>52.1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0.1071</td>\n",
       "      <td>228.22</td>\n",
       "      <td>11.082143</td>\n",
       "      <td>14.29</td>\n",
       "      <td>2760.000000</td>\n",
       "      <td>33623</td>\n",
       "      <td>76.7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0.1357</td>\n",
       "      <td>366.86</td>\n",
       "      <td>10.373491</td>\n",
       "      <td>11.63</td>\n",
       "      <td>4710.000000</td>\n",
       "      <td>3511</td>\n",
       "      <td>25.6</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0.1008</td>\n",
       "      <td>162.34</td>\n",
       "      <td>11.350407</td>\n",
       "      <td>8.10</td>\n",
       "      <td>2699.958333</td>\n",
       "      <td>33667</td>\n",
       "      <td>73.2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0.1426</td>\n",
       "      <td>102.92</td>\n",
       "      <td>11.299732</td>\n",
       "      <td>14.97</td>\n",
       "      <td>4066.000000</td>\n",
       "      <td>4740</td>\n",
       "      <td>39.5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   credit.policy  int.rate  installment  log.annual.inc    dti  \\\n",
       "0              1    0.1189       829.10       11.350407  19.48   \n",
       "1              1    0.1071       228.22       11.082143  14.29   \n",
       "2              1    0.1357       366.86       10.373491  11.63   \n",
       "3              1    0.1008       162.34       11.350407   8.10   \n",
       "4              1    0.1426       102.92       11.299732  14.97   \n",
       "\n",
       "   days.with.cr.line  revol.bal  revol.util  inq.last.6mths  delinq.2yrs  \\\n",
       "0        5639.958333      28854        52.1               0            0   \n",
       "1        2760.000000      33623        76.7               0            0   \n",
       "2        4710.000000       3511        25.6               1            0   \n",
       "3        2699.958333      33667        73.2               1            0   \n",
       "4        4066.000000       4740        39.5               0            1   \n",
       "\n",
       "   pub.rec  not.fully.paid  purpose_all_other  purpose_credit_card  \\\n",
       "0        0               0                  0                    0   \n",
       "1        0               0                  0                    1   \n",
       "2        0               0                  0                    0   \n",
       "3        0               0                  0                    0   \n",
       "4        0               0                  0                    1   \n",
       "\n",
       "   purpose_debt_consolidation  purpose_educational  purpose_home_improvement  \\\n",
       "0                           1                    0                         0   \n",
       "1                           0                    0                         0   \n",
       "2                           1                    0                         0   \n",
       "3                           1                    0                         0   \n",
       "4                           0                    0                         0   \n",
       "\n",
       "   purpose_major_purchase  purpose_small_business  \n",
       "0                       0                       0  \n",
       "1                       0                       0  \n",
       "2                       0                       0  \n",
       "3                       0                       0  \n",
       "4                       0                       0  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# removing features with (+-)0.7 correlation -> fico and int.rate are highly correlated so removing fico\n",
    "raw_df = raw_df.drop(['fico'],axis=1)\n",
    "raw_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((9578, 18), (9578,), (9578, 19), 18)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = raw_df.shape[1]-1\n",
    "x = raw_df.drop(['credit.policy'],axis=1).values\n",
    "y = raw_df['credit.policy'].values\n",
    "x.shape,y.shape,raw_df.shape,cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.13931753,  2.46309947,  0.68038804, ..., -0.2651173 ,\n",
       "        -0.21864717, -0.26285458],\n",
       "       [-0.57886837, -0.43885443,  0.2440308 , ..., -0.2651173 ,\n",
       "        -0.21864717, -0.26285458],\n",
       "       [ 0.48648368,  0.23070836, -0.90865897, ..., -0.2651173 ,\n",
       "        -0.21864717, -0.26285458],\n",
       "       ...,\n",
       "       [-0.57886837, -1.06867038, -0.54569448, ..., -0.2651173 ,\n",
       "        -0.21864717, -0.26285458],\n",
       "       [ 1.39166043,  0.1569135 , -0.18272998, ...,  3.77191529,\n",
       "        -0.21864717, -0.26285458],\n",
       "       [ 0.61685894,  2.58060136,  0.54059439, ..., -0.2651173 ,\n",
       "        -0.21864717, -0.26285458]])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "sc = StandardScaler()\n",
    "x_t = sc.fit_transform(x)\n",
    "x_t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import BatchNormalization\n",
    "from keras.optimizers import SGD, Adam\n",
    "from keras.layers import Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(Dense(20, input_dim=cols, activation='relu', kernel_initializer='he_uniform'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Dense(30,kernel_initializer='he_uniform', activation='relu'))\n",
    "model.add(Dense(20,kernel_initializer='he_uniform', activation='relu'))\n",
    "model.add(Dense(1, activation='sigmoid'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "epoch=50\n",
    "lr = 0.001\n",
    "decay_rate = lr / epoch\n",
    "momentum = 0.9\n",
    "sgd = SGD(lr=lr, momentum=momentum, decay=decay_rate, nesterov=False)\n",
    "model.compile(loss='binary_crossentropy', optimizer=sgd, metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((6704, 18), (2874, 18))"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "x_train, x_test, y_train, y_test = train_test_split(x_t, y, test_size=0.30, random_state=1)\n",
    "x_train.shape, x_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 6704 samples, validate on 2874 samples\n",
      "Epoch 1/50\n",
      "6704/6704 [==============================] - 0s 57us/step - loss: 0.7355 - accuracy: 0.5837 - val_loss: 0.5659 - val_accuracy: 0.7683\n",
      "Epoch 2/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.5299 - accuracy: 0.7874 - val_loss: 0.4982 - val_accuracy: 0.8072\n",
      "Epoch 3/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.4807 - accuracy: 0.8112 - val_loss: 0.4592 - val_accuracy: 0.8267\n",
      "Epoch 4/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.4492 - accuracy: 0.8234 - val_loss: 0.4309 - val_accuracy: 0.8340\n",
      "Epoch 5/50\n",
      "6704/6704 [==============================] - 0s 18us/step - loss: 0.4265 - accuracy: 0.8374 - val_loss: 0.4087 - val_accuracy: 0.8473\n",
      "Epoch 6/50\n",
      "6704/6704 [==============================] - 0s 19us/step - loss: 0.4069 - accuracy: 0.8408 - val_loss: 0.3908 - val_accuracy: 0.8546\n",
      "Epoch 7/50\n",
      "6704/6704 [==============================] - 0s 18us/step - loss: 0.3934 - accuracy: 0.8511 - val_loss: 0.3773 - val_accuracy: 0.8580\n",
      "Epoch 8/50\n",
      "6704/6704 [==============================] - 0s 17us/step - loss: 0.3822 - accuracy: 0.8550 - val_loss: 0.3649 - val_accuracy: 0.8615\n",
      "Epoch 9/50\n",
      "6704/6704 [==============================] - 0s 17us/step - loss: 0.3706 - accuracy: 0.8556 - val_loss: 0.3557 - val_accuracy: 0.8612\n",
      "Epoch 10/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.3631 - accuracy: 0.8572 - val_loss: 0.3460 - val_accuracy: 0.8667\n",
      "Epoch 11/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.3525 - accuracy: 0.8646 - val_loss: 0.3371 - val_accuracy: 0.8695\n",
      "Epoch 12/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.3462 - accuracy: 0.8661 - val_loss: 0.3300 - val_accuracy: 0.8737\n",
      "Epoch 13/50\n",
      "6704/6704 [==============================] - 0s 18us/step - loss: 0.3400 - accuracy: 0.8707 - val_loss: 0.3230 - val_accuracy: 0.8772\n",
      "Epoch 14/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.3325 - accuracy: 0.8705 - val_loss: 0.3171 - val_accuracy: 0.8807\n",
      "Epoch 15/50\n",
      "6704/6704 [==============================] - 0s 17us/step - loss: 0.3270 - accuracy: 0.8740 - val_loss: 0.3117 - val_accuracy: 0.8827\n",
      "Epoch 16/50\n",
      "6704/6704 [==============================] - 0s 19us/step - loss: 0.3249 - accuracy: 0.8719 - val_loss: 0.3067 - val_accuracy: 0.8866\n",
      "Epoch 17/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.3173 - accuracy: 0.8787 - val_loss: 0.3028 - val_accuracy: 0.8869\n",
      "Epoch 18/50\n",
      "6704/6704 [==============================] - 0s 17us/step - loss: 0.3159 - accuracy: 0.8747 - val_loss: 0.2994 - val_accuracy: 0.8883\n",
      "Epoch 19/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.3121 - accuracy: 0.8802 - val_loss: 0.2963 - val_accuracy: 0.8887\n",
      "Epoch 20/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.3102 - accuracy: 0.8798 - val_loss: 0.2942 - val_accuracy: 0.8897\n",
      "Epoch 21/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.3050 - accuracy: 0.8828 - val_loss: 0.2920 - val_accuracy: 0.8897\n",
      "Epoch 22/50\n",
      "6704/6704 [==============================] - 0s 18us/step - loss: 0.3022 - accuracy: 0.8814 - val_loss: 0.2897 - val_accuracy: 0.8918\n",
      "Epoch 23/50\n",
      "6704/6704 [==============================] - 0s 17us/step - loss: 0.2995 - accuracy: 0.8853 - val_loss: 0.2875 - val_accuracy: 0.8928\n",
      "Epoch 24/50\n",
      "6704/6704 [==============================] - 0s 24us/step - loss: 0.3001 - accuracy: 0.8835 - val_loss: 0.2860 - val_accuracy: 0.8921\n",
      "Epoch 25/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2953 - accuracy: 0.8841 - val_loss: 0.2855 - val_accuracy: 0.8904\n",
      "Epoch 26/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2950 - accuracy: 0.8857 - val_loss: 0.2834 - val_accuracy: 0.8914\n",
      "Epoch 27/50\n",
      "6704/6704 [==============================] - 0s 17us/step - loss: 0.2958 - accuracy: 0.8817 - val_loss: 0.2815 - val_accuracy: 0.8932\n",
      "Epoch 28/50\n",
      "6704/6704 [==============================] - 0s 17us/step - loss: 0.2920 - accuracy: 0.8862 - val_loss: 0.2801 - val_accuracy: 0.8918\n",
      "Epoch 29/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2933 - accuracy: 0.8860 - val_loss: 0.2790 - val_accuracy: 0.8928\n",
      "Epoch 30/50\n",
      "6704/6704 [==============================] - 0s 18us/step - loss: 0.2888 - accuracy: 0.8856 - val_loss: 0.2776 - val_accuracy: 0.8932\n",
      "Epoch 31/50\n",
      "6704/6704 [==============================] - 0s 17us/step - loss: 0.2892 - accuracy: 0.8851 - val_loss: 0.2766 - val_accuracy: 0.8918\n",
      "Epoch 32/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2870 - accuracy: 0.8889 - val_loss: 0.2764 - val_accuracy: 0.8935\n",
      "Epoch 33/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2863 - accuracy: 0.8859 - val_loss: 0.2763 - val_accuracy: 0.8928\n",
      "Epoch 34/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2865 - accuracy: 0.8853 - val_loss: 0.2750 - val_accuracy: 0.8935\n",
      "Epoch 35/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2859 - accuracy: 0.8865 - val_loss: 0.2741 - val_accuracy: 0.8935\n",
      "Epoch 36/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2810 - accuracy: 0.8892 - val_loss: 0.2731 - val_accuracy: 0.8932\n",
      "Epoch 37/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2826 - accuracy: 0.8881 - val_loss: 0.2721 - val_accuracy: 0.8928\n",
      "Epoch 38/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2836 - accuracy: 0.8881 - val_loss: 0.2710 - val_accuracy: 0.8939\n",
      "Epoch 39/50\n",
      "6704/6704 [==============================] - 0s 18us/step - loss: 0.2789 - accuracy: 0.8893 - val_loss: 0.2699 - val_accuracy: 0.8949\n",
      "Epoch 40/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2764 - accuracy: 0.8883 - val_loss: 0.2689 - val_accuracy: 0.8960\n",
      "Epoch 41/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2769 - accuracy: 0.8892 - val_loss: 0.2683 - val_accuracy: 0.8956\n",
      "Epoch 42/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2741 - accuracy: 0.8925 - val_loss: 0.2674 - val_accuracy: 0.8967\n",
      "Epoch 43/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2751 - accuracy: 0.8920 - val_loss: 0.2669 - val_accuracy: 0.8949\n",
      "Epoch 44/50\n",
      "6704/6704 [==============================] - 0s 16us/step - loss: 0.2720 - accuracy: 0.8939 - val_loss: 0.2664 - val_accuracy: 0.8949\n",
      "Epoch 45/50\n",
      "6704/6704 [==============================] - 0s 23us/step - loss: 0.2736 - accuracy: 0.8925 - val_loss: 0.2659 - val_accuracy: 0.8953\n",
      "Epoch 46/50\n",
      "6704/6704 [==============================] - 0s 35us/step - loss: 0.2716 - accuracy: 0.8929 - val_loss: 0.2652 - val_accuracy: 0.8974\n",
      "Epoch 47/50\n",
      "6704/6704 [==============================] - 0s 32us/step - loss: 0.2723 - accuracy: 0.8917 - val_loss: 0.2645 - val_accuracy: 0.8960\n",
      "Epoch 48/50\n",
      "6704/6704 [==============================] - 0s 27us/step - loss: 0.2670 - accuracy: 0.8939 - val_loss: 0.2642 - val_accuracy: 0.8977\n",
      "Epoch 49/50\n",
      "6704/6704 [==============================] - 0s 31us/step - loss: 0.2693 - accuracy: 0.8939 - val_loss: 0.2634 - val_accuracy: 0.8984\n",
      "Epoch 50/50\n",
      "6704/6704 [==============================] - 0s 32us/step - loss: 0.2677 - accuracy: 0.8941 - val_loss: 0.2631 - val_accuracy: 0.8987\n"
     ]
    }
   ],
   "source": [
    "batch = int(x_train.shape[0]/50)\n",
    "nn_model = model.fit(x_train, y_train,validation_data=(x_test, y_test), epochs=epoch,batch_size = batch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train: 0.898, Test: 0.899\n"
     ]
    }
   ],
   "source": [
    "_, train_acc = model.evaluate(x_train, y_train, verbose=0)\n",
    "_, test_acc = model.evaluate(x_test, y_test, verbose=0)\n",
    "print('Train: %.3f, Test: %.3f' % (train_acc, test_acc))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.plot(nn_model.history['accuracy'])\n",
    "plt.plot(nn_model.history['val_accuracy'])\n",
    "plt.title('Model accuracy')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(nn_model.history['loss']) \n",
    "plt.plot(nn_model.history['val_loss']) \n",
    "plt.title('Model loss') \n",
    "plt.ylabel('Loss') \n",
    "plt.xlabel('Epoch') \n",
    "plt.legend(['Train', 'Test'], loc='upper left') \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hyper parameter tunning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "from keras.wrappers.scikit_learn import KerasClassifier\n",
    "from sklearn.model_selection import GridSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 36 candidates, totalling 108 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=2)]: Using backend LokyBackend with 2 concurrent workers.\n",
      "[Parallel(n_jobs=2)]: Done   1 tasks      | elapsed:   15.6s\n",
      "[Parallel(n_jobs=2)]: Done   4 tasks      | elapsed:   20.8s\n",
      "[Parallel(n_jobs=2)]: Done   9 tasks      | elapsed:   36.7s\n",
      "[Parallel(n_jobs=2)]: Done  14 tasks      | elapsed:   51.5s\n",
      "[Parallel(n_jobs=2)]: Done  21 tasks      | elapsed:  1.4min\n",
      "[Parallel(n_jobs=2)]: Done  28 tasks      | elapsed:  1.9min\n",
      "[Parallel(n_jobs=2)]: Done  37 tasks      | elapsed:  2.8min\n",
      "[Parallel(n_jobs=2)]: Done  46 tasks      | elapsed:  3.2min\n",
      "[Parallel(n_jobs=2)]: Done  57 tasks      | elapsed:  4.1min\n",
      "[Parallel(n_jobs=2)]: Done  68 tasks      | elapsed:  5.2min\n",
      "[Parallel(n_jobs=2)]: Done  81 tasks      | elapsed:  6.1min\n",
      "[Parallel(n_jobs=2)]: Done  94 tasks      | elapsed:  6.9min\n",
      "[Parallel(n_jobs=2)]: Done 108 out of 108 | elapsed:  8.3min finished\n"
     ]
    }
   ],
   "source": [
    "s = x_train.shape[0]\n",
    "init_mode = ['glorot_normal', 'glorot_uniform', 'he_uniform']\n",
    "batch_size = [ int(s/20), int(s/50)]\n",
    "epochs = [75,100]\n",
    "optimizer = ['SGD', 'RMSprop', 'Adam']\n",
    "learn_rate = [0.001, 0.01]\n",
    "momentum = [0.0, 0.5, 0.9] \n",
    "activation = ['softplus', 'softmax', 'relu']\n",
    "dropout_rate = [0.2, 0.4, 0.6, 0.8, 0.9]\n",
    "from keras.constraints import maxnorm\n",
    "def create_model(init_mode='glorot_normal',activation='relu',dropout_rate=0.0, learn_rate=0.001,\n",
    "                neuron=15,optimizer='Adam'):\n",
    "    model = Sequential()\n",
    "    model.add(Dense(neuron, input_dim=cols, activation=activation, kernel_initializer=init_mode))\n",
    "    model.add(BatchNormalization())\n",
    "    model.add(Dense(neuron, activation=activation))\n",
    "    model.add(Dropout(dropout_rate))\n",
    "    model.add(Dense(neuron, activation=activation))\n",
    "    model.add(Dense(1, activation='sigmoid'))\n",
    "    optimizer = Adam(learning_rate=learn_rate)\n",
    "    model.compile(loss='binary_crossentropy', optimizer=optimizer, metrics=['accuracy'])\n",
    "    return model\n",
    "\n",
    "model_CV = KerasClassifier(build_fn=create_model, verbose=0,epochs=75, batch_size=134)\n",
    "param_grid = dict(init_mode=init_mode,batch_size=batch_size,epochs=epochs,activation=activation)\n",
    "grid_search = GridSearchCV(estimator=model_CV, param_grid=param_grid,cv=3,\n",
    "                           scoring='accuracy', verbose=10, n_jobs=2)\n",
    "result = grid_search.fit(x_train, y_train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Accuracy for 0.9197500896256553 using {'activation': 'softmax', 'batch_size': 134, 'epochs': 75, 'init_mode': 'glorot_uniform'}\n",
      "\n",
      " mean=0.9056, std=0.003544 using {'activation': 'softplus', 'batch_size': 335, 'epochs': 75, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9007, std=0.0003837 using {'activation': 'softplus', 'batch_size': 335, 'epochs': 75, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.9021, std=0.0005623 using {'activation': 'softplus', 'batch_size': 335, 'epochs': 75, 'init_mode': 'he_uniform'}\n",
      " mean=0.9017, std=0.002442 using {'activation': 'softplus', 'batch_size': 335, 'epochs': 100, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9021, std=0.0009007 using {'activation': 'softplus', 'batch_size': 335, 'epochs': 100, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.9023, std=0.003717 using {'activation': 'softplus', 'batch_size': 335, 'epochs': 100, 'init_mode': 'he_uniform'}\n",
      " mean=0.9083, std=0.00417 using {'activation': 'softplus', 'batch_size': 134, 'epochs': 75, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9081, std=0.004013 using {'activation': 'softplus', 'batch_size': 134, 'epochs': 75, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.9083, std=0.002512 using {'activation': 'softplus', 'batch_size': 134, 'epochs': 75, 'init_mode': 'he_uniform'}\n",
      " mean=0.9112, std=0.003675 using {'activation': 'softplus', 'batch_size': 134, 'epochs': 100, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9093, std=0.001372 using {'activation': 'softplus', 'batch_size': 134, 'epochs': 100, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.9047, std=0.001678 using {'activation': 'softplus', 'batch_size': 134, 'epochs': 100, 'init_mode': 'he_uniform'}\n",
      " mean=0.8971, std=0.001601 using {'activation': 'softmax', 'batch_size': 335, 'epochs': 75, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9038, std=0.001096 using {'activation': 'softmax', 'batch_size': 335, 'epochs': 75, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.9013, std=0.004018 using {'activation': 'softmax', 'batch_size': 335, 'epochs': 75, 'init_mode': 'he_uniform'}\n",
      " mean=0.9023, std=0.004975 using {'activation': 'softmax', 'batch_size': 335, 'epochs': 100, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9059, std=0.00174 using {'activation': 'softmax', 'batch_size': 335, 'epochs': 100, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.9013, std=0.008754 using {'activation': 'softmax', 'batch_size': 335, 'epochs': 100, 'init_mode': 'he_uniform'}\n",
      " mean=0.9065, std=0.005727 using {'activation': 'softmax', 'batch_size': 134, 'epochs': 75, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9198, std=0.003299 using {'activation': 'softmax', 'batch_size': 134, 'epochs': 75, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.9039, std=0.004102 using {'activation': 'softmax', 'batch_size': 134, 'epochs': 75, 'init_mode': 'he_uniform'}\n",
      " mean=0.9156, std=0.00697 using {'activation': 'softmax', 'batch_size': 134, 'epochs': 100, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9081, std=0.009411 using {'activation': 'softmax', 'batch_size': 134, 'epochs': 100, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.9139, std=0.0009363 using {'activation': 'softmax', 'batch_size': 134, 'epochs': 100, 'init_mode': 'he_uniform'}\n",
      " mean=0.9026, std=0.002359 using {'activation': 'relu', 'batch_size': 335, 'epochs': 75, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9068, std=0.002453 using {'activation': 'relu', 'batch_size': 335, 'epochs': 75, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.8987, std=0.002437 using {'activation': 'relu', 'batch_size': 335, 'epochs': 75, 'init_mode': 'he_uniform'}\n",
      " mean=0.9026, std=0.001045 using {'activation': 'relu', 'batch_size': 335, 'epochs': 100, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9021, std=0.003125 using {'activation': 'relu', 'batch_size': 335, 'epochs': 100, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.902, std=0.008861 using {'activation': 'relu', 'batch_size': 335, 'epochs': 100, 'init_mode': 'he_uniform'}\n",
      " mean=0.8998, std=0.0055 using {'activation': 'relu', 'batch_size': 134, 'epochs': 75, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9069, std=0.001671 using {'activation': 'relu', 'batch_size': 134, 'epochs': 75, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.9017, std=0.006831 using {'activation': 'relu', 'batch_size': 134, 'epochs': 75, 'init_mode': 'he_uniform'}\n",
      " mean=0.9062, std=0.003713 using {'activation': 'relu', 'batch_size': 134, 'epochs': 100, 'init_mode': 'glorot_normal'}\n",
      " mean=0.9072, std=0.00681 using {'activation': 'relu', 'batch_size': 134, 'epochs': 100, 'init_mode': 'glorot_uniform'}\n",
      " mean=0.9038, std=0.004562 using {'activation': 'relu', 'batch_size': 134, 'epochs': 100, 'init_mode': 'he_uniform'}\n"
     ]
    }
   ],
   "source": [
    "print(f'Best Accuracy for {result.best_score_} using {result.best_params_}\\n')\n",
    "means = result.cv_results_['mean_test_score']\n",
    "stds = result.cv_results_['std_test_score']\n",
    "params = result.cv_results_['params']\n",
    "for mean, stdev, param in zip(means, stds, params):\n",
    "    print(f' mean={mean:.4}, std={stdev:.4} using {param}')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 15 candidates, totalling 45 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=2)]: Using backend LokyBackend with 2 concurrent workers.\n",
      "[Parallel(n_jobs=2)]: Done   1 tasks      | elapsed:   53.6s\n",
      "[Parallel(n_jobs=2)]: Done   4 tasks      | elapsed:  1.5min\n",
      "[Parallel(n_jobs=2)]: Done   9 tasks      | elapsed:  3.4min\n",
      "[Parallel(n_jobs=2)]: Done  14 tasks      | elapsed:  4.5min\n",
      "[Parallel(n_jobs=2)]: Done  21 tasks      | elapsed:  6.8min\n",
      "[Parallel(n_jobs=2)]: Done  28 tasks      | elapsed:  8.7min\n",
      "[Parallel(n_jobs=2)]: Done  37 tasks      | elapsed: 11.5min\n",
      "[Parallel(n_jobs=2)]: Done  45 out of  45 | elapsed: 13.8min finished\n"
     ]
    }
   ],
   "source": [
    "optimizer = ['SGD', 'RMSprop', 'Adam']\n",
    "learn_rate = [0.001, 0.01]\n",
    "momentum = [0.0, 0.5, 0.9] \n",
    "dropout_rate = [0.2, 0.4, 0.6, 0.8, 0.9]\n",
    "# neurons = [10,15,20,25,30]\n",
    "from keras.constraints import maxnorm\n",
    "def create_model(init_mode='he_uniform',activation='relu',dropout_rate=0.0,\n",
    "                neuron=15,optimizer='Adam'):\n",
    "    model = Sequential()\n",
    "    model.add(Dense(neuron, input_dim=cols, activation=activation, kernel_initializer=init_mode))\n",
    "    model.add(BatchNormalization())\n",
    "    model.add(Dense(neuron, activation=activation))\n",
    "    model.add(Dropout(dropout_rate))\n",
    "    model.add(Dense(neuron, activation=activation))\n",
    "    model.add(Dense(1, activation='sigmoid'))\n",
    "    #optimizer = Adam(learning_rate=learn_rate)\n",
    "    model.compile(loss='binary_crossentropy', optimizer=optimizer, metrics=['accuracy'])\n",
    "    return model\n",
    "\n",
    "\n",
    "model_CV = KerasClassifier(build_fn=create_model, verbose=0,epochs=100, batch_size=30)\n",
    "param_grid = dict(optimizer=optimizer,dropout_rate=dropout_rate)\n",
    "grid_search = GridSearchCV(estimator=model_CV, param_grid=param_grid,cv=3,\n",
    "                           scoring='accuracy', verbose=10,n_jobs=2)\n",
    "result = grid_search.fit(x_train, y_train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Accuracy for 0.909756078021386 using {'dropout_rate': 0.2, 'optimizer': 'Adam'}\n",
      "\n",
      " mean=0.8999, std=0.003507 using {'dropout_rate': 0.2, 'optimizer': 'SGD'}\n",
      " mean=0.9078, std=0.002881 using {'dropout_rate': 0.2, 'optimizer': 'RMSprop'}\n",
      " mean=0.9098, std=0.005236 using {'dropout_rate': 0.2, 'optimizer': 'Adam'}\n",
      " mean=0.8987, std=0.003492 using {'dropout_rate': 0.4, 'optimizer': 'SGD'}\n",
      " mean=0.9056, std=0.004212 using {'dropout_rate': 0.4, 'optimizer': 'RMSprop'}\n",
      " mean=0.905, std=0.004727 using {'dropout_rate': 0.4, 'optimizer': 'Adam'}\n",
      " mean=0.8743, std=0.01885 using {'dropout_rate': 0.6, 'optimizer': 'SGD'}\n",
      " mean=0.8993, std=0.01252 using {'dropout_rate': 0.6, 'optimizer': 'RMSprop'}\n",
      " mean=0.8966, std=0.001888 using {'dropout_rate': 0.6, 'optimizer': 'Adam'}\n",
      " mean=0.8477, std=0.02888 using {'dropout_rate': 0.8, 'optimizer': 'SGD'}\n",
      " mean=0.8799, std=0.01789 using {'dropout_rate': 0.8, 'optimizer': 'RMSprop'}\n",
      " mean=0.8599, std=0.007645 using {'dropout_rate': 0.8, 'optimizer': 'Adam'}\n",
      " mean=0.8164, std=0.01314 using {'dropout_rate': 0.9, 'optimizer': 'SGD'}\n",
      " mean=0.8167, std=0.01436 using {'dropout_rate': 0.9, 'optimizer': 'RMSprop'}\n",
      " mean=0.8376, std=0.007484 using {'dropout_rate': 0.9, 'optimizer': 'Adam'}\n"
     ]
    }
   ],
   "source": [
    "print(f'Best Accuracy for {result.best_score_} using {result.best_params_}\\n')\n",
    "means = result.cv_results_['mean_test_score']\n",
    "stds = result.cv_results_['std_test_score']\n",
    "params = result.cv_results_['params']\n",
    "for mean, stdev, param in zip(means, stds, params):\n",
    "    print(f' mean={mean:.4}, std={stdev:.4} using {param}')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Final Model Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 6704 samples, validate on 2874 samples\n",
      "Epoch 1/20\n",
      "6704/6704 [==============================] - 1s 189us/step - loss: 0.5216 - accuracy: 0.7955 - val_loss: 0.4343 - val_accuracy: 0.8142\n",
      "Epoch 2/20\n",
      "6704/6704 [==============================] - 1s 102us/step - loss: 0.4390 - accuracy: 0.8031 - val_loss: 0.3590 - val_accuracy: 0.8264\n",
      "Epoch 3/20\n",
      "6704/6704 [==============================] - 1s 96us/step - loss: 0.3802 - accuracy: 0.8364 - val_loss: 0.3021 - val_accuracy: 0.8820\n",
      "Epoch 4/20\n",
      "6704/6704 [==============================] - 1s 88us/step - loss: 0.3403 - accuracy: 0.8628 - val_loss: 0.2817 - val_accuracy: 0.8894\n",
      "Epoch 5/20\n",
      "6704/6704 [==============================] - 1s 99us/step - loss: 0.3130 - accuracy: 0.8747 - val_loss: 0.2649 - val_accuracy: 0.8970\n",
      "Epoch 6/20\n",
      "6704/6704 [==============================] - 1s 96us/step - loss: 0.3036 - accuracy: 0.8793 - val_loss: 0.2654 - val_accuracy: 0.9033\n",
      "Epoch 7/20\n",
      "6704/6704 [==============================] - 1s 96us/step - loss: 0.2930 - accuracy: 0.8847 - val_loss: 0.2545 - val_accuracy: 0.9036\n",
      "Epoch 8/20\n",
      "6704/6704 [==============================] - 1s 97us/step - loss: 0.2836 - accuracy: 0.8828 - val_loss: 0.2505 - val_accuracy: 0.9085\n",
      "Epoch 9/20\n",
      "6704/6704 [==============================] - 1s 96us/step - loss: 0.2786 - accuracy: 0.8911 - val_loss: 0.2488 - val_accuracy: 0.9068\n",
      "Epoch 10/20\n",
      "6704/6704 [==============================] - 1s 94us/step - loss: 0.2818 - accuracy: 0.8859 - val_loss: 0.2464 - val_accuracy: 0.9071\n",
      "Epoch 11/20\n",
      "6704/6704 [==============================] - 1s 95us/step - loss: 0.2800 - accuracy: 0.8880 - val_loss: 0.2488 - val_accuracy: 0.9099\n",
      "Epoch 12/20\n",
      "6704/6704 [==============================] - 1s 130us/step - loss: 0.2719 - accuracy: 0.8895 - val_loss: 0.2530 - val_accuracy: 0.9099\n",
      "Epoch 13/20\n",
      "6704/6704 [==============================] - 1s 102us/step - loss: 0.2705 - accuracy: 0.8926 - val_loss: 0.2498 - val_accuracy: 0.9071\n",
      "Epoch 14/20\n",
      "6704/6704 [==============================] - 1s 95us/step - loss: 0.2609 - accuracy: 0.8926 - val_loss: 0.2453 - val_accuracy: 0.9120\n",
      "Epoch 15/20\n",
      "6704/6704 [==============================] - 1s 93us/step - loss: 0.2644 - accuracy: 0.8923 - val_loss: 0.2430 - val_accuracy: 0.9113\n",
      "Epoch 16/20\n",
      "6704/6704 [==============================] - 1s 96us/step - loss: 0.2553 - accuracy: 0.8945 - val_loss: 0.2367 - val_accuracy: 0.9113\n",
      "Epoch 17/20\n",
      "6704/6704 [==============================] - 1s 97us/step - loss: 0.2614 - accuracy: 0.8913 - val_loss: 0.2448 - val_accuracy: 0.9081\n",
      "Epoch 18/20\n",
      "6704/6704 [==============================] - 1s 95us/step - loss: 0.2545 - accuracy: 0.8944 - val_loss: 0.2392 - val_accuracy: 0.9092\n",
      "Epoch 19/20\n",
      "6704/6704 [==============================] - 1s 94us/step - loss: 0.2514 - accuracy: 0.8939 - val_loss: 0.2464 - val_accuracy: 0.9085\n",
      "Epoch 20/20\n",
      "6704/6704 [==============================] - 1s 95us/step - loss: 0.2517 - accuracy: 0.9010 - val_loss: 0.2326 - val_accuracy: 0.9120\n",
      "Train: 0.918, Test: 0.912\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dr = 0.2\n",
    "model = Sequential()\n",
    "model.add(Dense(20, input_dim=cols, activation='relu', kernel_initializer='he_uniform'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Dense(30,kernel_initializer='he_uniform', activation='relu'))\n",
    "model.add(Dropout(dr))\n",
    "model.add(Dense(20,kernel_initializer='he_uniform', activation='relu'))\n",
    "model.add(Dropout(dr))\n",
    "model.add(Dense(20,kernel_initializer='he_uniform', activation='relu'))\n",
    "model.add(Dropout(dr))\n",
    "model.add(Dense(20,kernel_initializer='he_uniform', activation='relu'))\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "epoch=20\n",
    "lr = 0.001\n",
    "decay_rate = lr / epoch\n",
    "momentum = 0.9\n",
    "optimizer = Adam(learning_rate=lr)\n",
    "model.compile(loss='binary_crossentropy', optimizer=optimizer, metrics=['accuracy'])\n",
    "batch = 30 #int(x_train.shape[0]/50)\n",
    "nn_model = model.fit(x_train, y_train,validation_data=(x_test, y_test), epochs=epoch,batch_size = batch)\n",
    "_, train_acc = model.evaluate(x_train, y_train, verbose=0)\n",
    "_, test_acc = model.evaluate(x_test, y_test, verbose=0)\n",
    "print('Train: %.3f, Test: %.3f' % (train_acc, test_acc))\n",
    "plt.plot(nn_model.history['accuracy'])\n",
    "plt.plot(nn_model.history['val_accuracy'])\n",
    "plt.title('Model accuracy')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Test'], loc='upper left')\n",
    "plt.show()\n",
    "plt.plot(nn_model.history['loss']) \n",
    "plt.plot(nn_model.history['val_loss']) \n",
    "plt.title('Model loss') \n",
    "plt.ylabel('Loss') \n",
    "plt.xlabel('Epoch') \n",
    "plt.legend(['Train', 'Test'], loc='upper left') \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Accuracy : Train: 0.918, Test: 0.912"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}