Deep Learning + Sampling.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import keras\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "np.random.seed(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv('creditcard.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data exploration"
   ]
  },
  {
   "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>Time</th>\n",
       "      <th>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>V9</th>\n",
       "      <th>...</th>\n",
       "      <th>V21</th>\n",
       "      <th>V22</th>\n",
       "      <th>V23</th>\n",
       "      <th>V24</th>\n",
       "      <th>V25</th>\n",
       "      <th>V26</th>\n",
       "      <th>V27</th>\n",
       "      <th>V28</th>\n",
       "      <th>Amount</th>\n",
       "      <th>Class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.359807</td>\n",
       "      <td>-0.072781</td>\n",
       "      <td>2.536347</td>\n",
       "      <td>1.378155</td>\n",
       "      <td>-0.338321</td>\n",
       "      <td>0.462388</td>\n",
       "      <td>0.239599</td>\n",
       "      <td>0.098698</td>\n",
       "      <td>0.363787</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.018307</td>\n",
       "      <td>0.277838</td>\n",
       "      <td>-0.110474</td>\n",
       "      <td>0.066928</td>\n",
       "      <td>0.128539</td>\n",
       "      <td>-0.189115</td>\n",
       "      <td>0.133558</td>\n",
       "      <td>-0.021053</td>\n",
       "      <td>149.62</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.191857</td>\n",
       "      <td>0.266151</td>\n",
       "      <td>0.166480</td>\n",
       "      <td>0.448154</td>\n",
       "      <td>0.060018</td>\n",
       "      <td>-0.082361</td>\n",
       "      <td>-0.078803</td>\n",
       "      <td>0.085102</td>\n",
       "      <td>-0.255425</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.225775</td>\n",
       "      <td>-0.638672</td>\n",
       "      <td>0.101288</td>\n",
       "      <td>-0.339846</td>\n",
       "      <td>0.167170</td>\n",
       "      <td>0.125895</td>\n",
       "      <td>-0.008983</td>\n",
       "      <td>0.014724</td>\n",
       "      <td>2.69</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.358354</td>\n",
       "      <td>-1.340163</td>\n",
       "      <td>1.773209</td>\n",
       "      <td>0.379780</td>\n",
       "      <td>-0.503198</td>\n",
       "      <td>1.800499</td>\n",
       "      <td>0.791461</td>\n",
       "      <td>0.247676</td>\n",
       "      <td>-1.514654</td>\n",
       "      <td>...</td>\n",
       "      <td>0.247998</td>\n",
       "      <td>0.771679</td>\n",
       "      <td>0.909412</td>\n",
       "      <td>-0.689281</td>\n",
       "      <td>-0.327642</td>\n",
       "      <td>-0.139097</td>\n",
       "      <td>-0.055353</td>\n",
       "      <td>-0.059752</td>\n",
       "      <td>378.66</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.966272</td>\n",
       "      <td>-0.185226</td>\n",
       "      <td>1.792993</td>\n",
       "      <td>-0.863291</td>\n",
       "      <td>-0.010309</td>\n",
       "      <td>1.247203</td>\n",
       "      <td>0.237609</td>\n",
       "      <td>0.377436</td>\n",
       "      <td>-1.387024</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.108300</td>\n",
       "      <td>0.005274</td>\n",
       "      <td>-0.190321</td>\n",
       "      <td>-1.175575</td>\n",
       "      <td>0.647376</td>\n",
       "      <td>-0.221929</td>\n",
       "      <td>0.062723</td>\n",
       "      <td>0.061458</td>\n",
       "      <td>123.50</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2.0</td>\n",
       "      <td>-1.158233</td>\n",
       "      <td>0.877737</td>\n",
       "      <td>1.548718</td>\n",
       "      <td>0.403034</td>\n",
       "      <td>-0.407193</td>\n",
       "      <td>0.095921</td>\n",
       "      <td>0.592941</td>\n",
       "      <td>-0.270533</td>\n",
       "      <td>0.817739</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.009431</td>\n",
       "      <td>0.798278</td>\n",
       "      <td>-0.137458</td>\n",
       "      <td>0.141267</td>\n",
       "      <td>-0.206010</td>\n",
       "      <td>0.502292</td>\n",
       "      <td>0.219422</td>\n",
       "      <td>0.215153</td>\n",
       "      <td>69.99</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Time        V1        V2        V3        V4        V5        V6        V7  \\\n",
       "0   0.0 -1.359807 -0.072781  2.536347  1.378155 -0.338321  0.462388  0.239599   \n",
       "1   0.0  1.191857  0.266151  0.166480  0.448154  0.060018 -0.082361 -0.078803   \n",
       "2   1.0 -1.358354 -1.340163  1.773209  0.379780 -0.503198  1.800499  0.791461   \n",
       "3   1.0 -0.966272 -0.185226  1.792993 -0.863291 -0.010309  1.247203  0.237609   \n",
       "4   2.0 -1.158233  0.877737  1.548718  0.403034 -0.407193  0.095921  0.592941   \n",
       "\n",
       "         V8        V9  ...       V21       V22       V23       V24       V25  \\\n",
       "0  0.098698  0.363787  ... -0.018307  0.277838 -0.110474  0.066928  0.128539   \n",
       "1  0.085102 -0.255425  ... -0.225775 -0.638672  0.101288 -0.339846  0.167170   \n",
       "2  0.247676 -1.514654  ...  0.247998  0.771679  0.909412 -0.689281 -0.327642   \n",
       "3  0.377436 -1.387024  ... -0.108300  0.005274 -0.190321 -1.175575  0.647376   \n",
       "4 -0.270533  0.817739  ... -0.009431  0.798278 -0.137458  0.141267 -0.206010   \n",
       "\n",
       "        V26       V27       V28  Amount  Class  \n",
       "0 -0.189115  0.133558 -0.021053  149.62      0  \n",
       "1  0.125895 -0.008983  0.014724    2.69      0  \n",
       "2 -0.139097 -0.055353 -0.059752  378.66      0  \n",
       "3 -0.221929  0.062723  0.061458  123.50      0  \n",
       "4  0.502292  0.219422  0.215153   69.99      0  \n",
       "\n",
       "[5 rows x 31 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "284807"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>Time</th>\n",
       "      <th>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>V9</th>\n",
       "      <th>...</th>\n",
       "      <th>V21</th>\n",
       "      <th>V22</th>\n",
       "      <th>V23</th>\n",
       "      <th>V24</th>\n",
       "      <th>V25</th>\n",
       "      <th>V26</th>\n",
       "      <th>V27</th>\n",
       "      <th>V28</th>\n",
       "      <th>Amount</th>\n",
       "      <th>Class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>284807.000000</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>...</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>2.848070e+05</td>\n",
       "      <td>284807.000000</td>\n",
       "      <td>284807.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>94813.859575</td>\n",
       "      <td>1.165980e-15</td>\n",
       "      <td>3.416908e-16</td>\n",
       "      <td>-1.373150e-15</td>\n",
       "      <td>2.086869e-15</td>\n",
       "      <td>9.604066e-16</td>\n",
       "      <td>1.490107e-15</td>\n",
       "      <td>-5.556467e-16</td>\n",
       "      <td>1.177556e-16</td>\n",
       "      <td>-2.406455e-15</td>\n",
       "      <td>...</td>\n",
       "      <td>1.656562e-16</td>\n",
       "      <td>-3.444850e-16</td>\n",
       "      <td>2.578648e-16</td>\n",
       "      <td>4.471968e-15</td>\n",
       "      <td>5.340915e-16</td>\n",
       "      <td>1.687098e-15</td>\n",
       "      <td>-3.666453e-16</td>\n",
       "      <td>-1.220404e-16</td>\n",
       "      <td>88.349619</td>\n",
       "      <td>0.001727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>47488.145955</td>\n",
       "      <td>1.958696e+00</td>\n",
       "      <td>1.651309e+00</td>\n",
       "      <td>1.516255e+00</td>\n",
       "      <td>1.415869e+00</td>\n",
       "      <td>1.380247e+00</td>\n",
       "      <td>1.332271e+00</td>\n",
       "      <td>1.237094e+00</td>\n",
       "      <td>1.194353e+00</td>\n",
       "      <td>1.098632e+00</td>\n",
       "      <td>...</td>\n",
       "      <td>7.345240e-01</td>\n",
       "      <td>7.257016e-01</td>\n",
       "      <td>6.244603e-01</td>\n",
       "      <td>6.056471e-01</td>\n",
       "      <td>5.212781e-01</td>\n",
       "      <td>4.822270e-01</td>\n",
       "      <td>4.036325e-01</td>\n",
       "      <td>3.300833e-01</td>\n",
       "      <td>250.120109</td>\n",
       "      <td>0.041527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-5.640751e+01</td>\n",
       "      <td>-7.271573e+01</td>\n",
       "      <td>-4.832559e+01</td>\n",
       "      <td>-5.683171e+00</td>\n",
       "      <td>-1.137433e+02</td>\n",
       "      <td>-2.616051e+01</td>\n",
       "      <td>-4.355724e+01</td>\n",
       "      <td>-7.321672e+01</td>\n",
       "      <td>-1.343407e+01</td>\n",
       "      <td>...</td>\n",
       "      <td>-3.483038e+01</td>\n",
       "      <td>-1.093314e+01</td>\n",
       "      <td>-4.480774e+01</td>\n",
       "      <td>-2.836627e+00</td>\n",
       "      <td>-1.029540e+01</td>\n",
       "      <td>-2.604551e+00</td>\n",
       "      <td>-2.256568e+01</td>\n",
       "      <td>-1.543008e+01</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>54201.500000</td>\n",
       "      <td>-9.203734e-01</td>\n",
       "      <td>-5.985499e-01</td>\n",
       "      <td>-8.903648e-01</td>\n",
       "      <td>-8.486401e-01</td>\n",
       "      <td>-6.915971e-01</td>\n",
       "      <td>-7.682956e-01</td>\n",
       "      <td>-5.540759e-01</td>\n",
       "      <td>-2.086297e-01</td>\n",
       "      <td>-6.430976e-01</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.283949e-01</td>\n",
       "      <td>-5.423504e-01</td>\n",
       "      <td>-1.618463e-01</td>\n",
       "      <td>-3.545861e-01</td>\n",
       "      <td>-3.171451e-01</td>\n",
       "      <td>-3.269839e-01</td>\n",
       "      <td>-7.083953e-02</td>\n",
       "      <td>-5.295979e-02</td>\n",
       "      <td>5.600000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>84692.000000</td>\n",
       "      <td>1.810880e-02</td>\n",
       "      <td>6.548556e-02</td>\n",
       "      <td>1.798463e-01</td>\n",
       "      <td>-1.984653e-02</td>\n",
       "      <td>-5.433583e-02</td>\n",
       "      <td>-2.741871e-01</td>\n",
       "      <td>4.010308e-02</td>\n",
       "      <td>2.235804e-02</td>\n",
       "      <td>-5.142873e-02</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.945017e-02</td>\n",
       "      <td>6.781943e-03</td>\n",
       "      <td>-1.119293e-02</td>\n",
       "      <td>4.097606e-02</td>\n",
       "      <td>1.659350e-02</td>\n",
       "      <td>-5.213911e-02</td>\n",
       "      <td>1.342146e-03</td>\n",
       "      <td>1.124383e-02</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>139320.500000</td>\n",
       "      <td>1.315642e+00</td>\n",
       "      <td>8.037239e-01</td>\n",
       "      <td>1.027196e+00</td>\n",
       "      <td>7.433413e-01</td>\n",
       "      <td>6.119264e-01</td>\n",
       "      <td>3.985649e-01</td>\n",
       "      <td>5.704361e-01</td>\n",
       "      <td>3.273459e-01</td>\n",
       "      <td>5.971390e-01</td>\n",
       "      <td>...</td>\n",
       "      <td>1.863772e-01</td>\n",
       "      <td>5.285536e-01</td>\n",
       "      <td>1.476421e-01</td>\n",
       "      <td>4.395266e-01</td>\n",
       "      <td>3.507156e-01</td>\n",
       "      <td>2.409522e-01</td>\n",
       "      <td>9.104512e-02</td>\n",
       "      <td>7.827995e-02</td>\n",
       "      <td>77.165000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>172792.000000</td>\n",
       "      <td>2.454930e+00</td>\n",
       "      <td>2.205773e+01</td>\n",
       "      <td>9.382558e+00</td>\n",
       "      <td>1.687534e+01</td>\n",
       "      <td>3.480167e+01</td>\n",
       "      <td>7.330163e+01</td>\n",
       "      <td>1.205895e+02</td>\n",
       "      <td>2.000721e+01</td>\n",
       "      <td>1.559499e+01</td>\n",
       "      <td>...</td>\n",
       "      <td>2.720284e+01</td>\n",
       "      <td>1.050309e+01</td>\n",
       "      <td>2.252841e+01</td>\n",
       "      <td>4.584549e+00</td>\n",
       "      <td>7.519589e+00</td>\n",
       "      <td>3.517346e+00</td>\n",
       "      <td>3.161220e+01</td>\n",
       "      <td>3.384781e+01</td>\n",
       "      <td>25691.160000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                Time            V1            V2            V3            V4  \\\n",
       "count  284807.000000  2.848070e+05  2.848070e+05  2.848070e+05  2.848070e+05   \n",
       "mean    94813.859575  1.165980e-15  3.416908e-16 -1.373150e-15  2.086869e-15   \n",
       "std     47488.145955  1.958696e+00  1.651309e+00  1.516255e+00  1.415869e+00   \n",
       "min         0.000000 -5.640751e+01 -7.271573e+01 -4.832559e+01 -5.683171e+00   \n",
       "25%     54201.500000 -9.203734e-01 -5.985499e-01 -8.903648e-01 -8.486401e-01   \n",
       "50%     84692.000000  1.810880e-02  6.548556e-02  1.798463e-01 -1.984653e-02   \n",
       "75%    139320.500000  1.315642e+00  8.037239e-01  1.027196e+00  7.433413e-01   \n",
       "max    172792.000000  2.454930e+00  2.205773e+01  9.382558e+00  1.687534e+01   \n",
       "\n",
       "                 V5            V6            V7            V8            V9  \\\n",
       "count  2.848070e+05  2.848070e+05  2.848070e+05  2.848070e+05  2.848070e+05   \n",
       "mean   9.604066e-16  1.490107e-15 -5.556467e-16  1.177556e-16 -2.406455e-15   \n",
       "std    1.380247e+00  1.332271e+00  1.237094e+00  1.194353e+00  1.098632e+00   \n",
       "min   -1.137433e+02 -2.616051e+01 -4.355724e+01 -7.321672e+01 -1.343407e+01   \n",
       "25%   -6.915971e-01 -7.682956e-01 -5.540759e-01 -2.086297e-01 -6.430976e-01   \n",
       "50%   -5.433583e-02 -2.741871e-01  4.010308e-02  2.235804e-02 -5.142873e-02   \n",
       "75%    6.119264e-01  3.985649e-01  5.704361e-01  3.273459e-01  5.971390e-01   \n",
       "max    3.480167e+01  7.330163e+01  1.205895e+02  2.000721e+01  1.559499e+01   \n",
       "\n",
       "       ...           V21           V22           V23           V24  \\\n",
       "count  ...  2.848070e+05  2.848070e+05  2.848070e+05  2.848070e+05   \n",
       "mean   ...  1.656562e-16 -3.444850e-16  2.578648e-16  4.471968e-15   \n",
       "std    ...  7.345240e-01  7.257016e-01  6.244603e-01  6.056471e-01   \n",
       "min    ... -3.483038e+01 -1.093314e+01 -4.480774e+01 -2.836627e+00   \n",
       "25%    ... -2.283949e-01 -5.423504e-01 -1.618463e-01 -3.545861e-01   \n",
       "50%    ... -2.945017e-02  6.781943e-03 -1.119293e-02  4.097606e-02   \n",
       "75%    ...  1.863772e-01  5.285536e-01  1.476421e-01  4.395266e-01   \n",
       "max    ...  2.720284e+01  1.050309e+01  2.252841e+01  4.584549e+00   \n",
       "\n",
       "                V25           V26           V27           V28         Amount  \\\n",
       "count  2.848070e+05  2.848070e+05  2.848070e+05  2.848070e+05  284807.000000   \n",
       "mean   5.340915e-16  1.687098e-15 -3.666453e-16 -1.220404e-16      88.349619   \n",
       "std    5.212781e-01  4.822270e-01  4.036325e-01  3.300833e-01     250.120109   \n",
       "min   -1.029540e+01 -2.604551e+00 -2.256568e+01 -1.543008e+01       0.000000   \n",
       "25%   -3.171451e-01 -3.269839e-01 -7.083953e-02 -5.295979e-02       5.600000   \n",
       "50%    1.659350e-02 -5.213911e-02  1.342146e-03  1.124383e-02      22.000000   \n",
       "75%    3.507156e-01  2.409522e-01  9.104512e-02  7.827995e-02      77.165000   \n",
       "max    7.519589e+00  3.517346e+00  3.161220e+01  3.384781e+01   25691.160000   \n",
       "\n",
       "               Class  \n",
       "count  284807.000000  \n",
       "mean        0.001727  \n",
       "std         0.041527  \n",
       "min         0.000000  \n",
       "25%         0.000000  \n",
       "50%         0.000000  \n",
       "75%         0.000000  \n",
       "max         1.000000  \n",
       "\n",
       "[8 rows x 31 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1440 with 31 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Histograms\n",
    "fig = plt.figure(figsize=(15, 20))\n",
    "plt.suptitle('Histograms of Numerical Columns', fontsize=20)\n",
    "for i in range(data.shape[1]):\n",
    "    plt.subplot(8, 4, i + 1)\n",
    "    f = plt.gca()\n",
    "    f.set_title(data.columns.values[i])\n",
    "\n",
    "    vals = np.size(data.iloc[:, i].unique())\n",
    "    if vals >= 100:\n",
    "        vals = 100                                    # limit our bins to 100 maximum\n",
    "    \n",
    "    plt.hist(data.iloc[:, i], bins=vals, color='#3F5D7D')\n",
    "plt.tight_layout(rect=[0, 0.03, 1, 0.95])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- we can observe a large prevalence of Class 0 (non fraudulent)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of fraudulent transactions = 492 or 172 per 100,000 transactions in the dataset\n"
     ]
    }
   ],
   "source": [
    "print('Number of fraudulent transactions = %d or %d per 100,000 transactions in the dataset'\n",
    "      %(len(data[data.Class==1]), len(data[data.Class==1])/len(data)*100000))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Linear Correlation with Response Variable (Note: Models like RandomForest are not linear)\n",
    "data2 = data.drop(columns = ['Class'])   # drop non numerical columns\n",
    "data2.corrwith(data.Class).plot.bar(\n",
    "        figsize = (20, 10), title = \"Correlation with Class Fraudulent or Not\", fontsize = 15,\n",
    "        rot = 45, grid = True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pre-processing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "data['normalizedAmount'] = StandardScaler().fit_transform(data['Amount'].values.reshape(-1,1))  # Normalize 'Amount' in [-1,+1] range\n",
    "data = data.drop(['Amount'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>Time</th>\n",
       "      <th>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>V9</th>\n",
       "      <th>...</th>\n",
       "      <th>V21</th>\n",
       "      <th>V22</th>\n",
       "      <th>V23</th>\n",
       "      <th>V24</th>\n",
       "      <th>V25</th>\n",
       "      <th>V26</th>\n",
       "      <th>V27</th>\n",
       "      <th>V28</th>\n",
       "      <th>Class</th>\n",
       "      <th>normalizedAmount</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.359807</td>\n",
       "      <td>-0.072781</td>\n",
       "      <td>2.536347</td>\n",
       "      <td>1.378155</td>\n",
       "      <td>-0.338321</td>\n",
       "      <td>0.462388</td>\n",
       "      <td>0.239599</td>\n",
       "      <td>0.098698</td>\n",
       "      <td>0.363787</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.018307</td>\n",
       "      <td>0.277838</td>\n",
       "      <td>-0.110474</td>\n",
       "      <td>0.066928</td>\n",
       "      <td>0.128539</td>\n",
       "      <td>-0.189115</td>\n",
       "      <td>0.133558</td>\n",
       "      <td>-0.021053</td>\n",
       "      <td>0</td>\n",
       "      <td>0.244964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.191857</td>\n",
       "      <td>0.266151</td>\n",
       "      <td>0.166480</td>\n",
       "      <td>0.448154</td>\n",
       "      <td>0.060018</td>\n",
       "      <td>-0.082361</td>\n",
       "      <td>-0.078803</td>\n",
       "      <td>0.085102</td>\n",
       "      <td>-0.255425</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.225775</td>\n",
       "      <td>-0.638672</td>\n",
       "      <td>0.101288</td>\n",
       "      <td>-0.339846</td>\n",
       "      <td>0.167170</td>\n",
       "      <td>0.125895</td>\n",
       "      <td>-0.008983</td>\n",
       "      <td>0.014724</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.342475</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.358354</td>\n",
       "      <td>-1.340163</td>\n",
       "      <td>1.773209</td>\n",
       "      <td>0.379780</td>\n",
       "      <td>-0.503198</td>\n",
       "      <td>1.800499</td>\n",
       "      <td>0.791461</td>\n",
       "      <td>0.247676</td>\n",
       "      <td>-1.514654</td>\n",
       "      <td>...</td>\n",
       "      <td>0.247998</td>\n",
       "      <td>0.771679</td>\n",
       "      <td>0.909412</td>\n",
       "      <td>-0.689281</td>\n",
       "      <td>-0.327642</td>\n",
       "      <td>-0.139097</td>\n",
       "      <td>-0.055353</td>\n",
       "      <td>-0.059752</td>\n",
       "      <td>0</td>\n",
       "      <td>1.160686</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.966272</td>\n",
       "      <td>-0.185226</td>\n",
       "      <td>1.792993</td>\n",
       "      <td>-0.863291</td>\n",
       "      <td>-0.010309</td>\n",
       "      <td>1.247203</td>\n",
       "      <td>0.237609</td>\n",
       "      <td>0.377436</td>\n",
       "      <td>-1.387024</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.108300</td>\n",
       "      <td>0.005274</td>\n",
       "      <td>-0.190321</td>\n",
       "      <td>-1.175575</td>\n",
       "      <td>0.647376</td>\n",
       "      <td>-0.221929</td>\n",
       "      <td>0.062723</td>\n",
       "      <td>0.061458</td>\n",
       "      <td>0</td>\n",
       "      <td>0.140534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2.0</td>\n",
       "      <td>-1.158233</td>\n",
       "      <td>0.877737</td>\n",
       "      <td>1.548718</td>\n",
       "      <td>0.403034</td>\n",
       "      <td>-0.407193</td>\n",
       "      <td>0.095921</td>\n",
       "      <td>0.592941</td>\n",
       "      <td>-0.270533</td>\n",
       "      <td>0.817739</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.009431</td>\n",
       "      <td>0.798278</td>\n",
       "      <td>-0.137458</td>\n",
       "      <td>0.141267</td>\n",
       "      <td>-0.206010</td>\n",
       "      <td>0.502292</td>\n",
       "      <td>0.219422</td>\n",
       "      <td>0.215153</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.073403</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Time        V1        V2        V3        V4        V5        V6        V7  \\\n",
       "0   0.0 -1.359807 -0.072781  2.536347  1.378155 -0.338321  0.462388  0.239599   \n",
       "1   0.0  1.191857  0.266151  0.166480  0.448154  0.060018 -0.082361 -0.078803   \n",
       "2   1.0 -1.358354 -1.340163  1.773209  0.379780 -0.503198  1.800499  0.791461   \n",
       "3   1.0 -0.966272 -0.185226  1.792993 -0.863291 -0.010309  1.247203  0.237609   \n",
       "4   2.0 -1.158233  0.877737  1.548718  0.403034 -0.407193  0.095921  0.592941   \n",
       "\n",
       "         V8        V9  ...       V21       V22       V23       V24       V25  \\\n",
       "0  0.098698  0.363787  ... -0.018307  0.277838 -0.110474  0.066928  0.128539   \n",
       "1  0.085102 -0.255425  ... -0.225775 -0.638672  0.101288 -0.339846  0.167170   \n",
       "2  0.247676 -1.514654  ...  0.247998  0.771679  0.909412 -0.689281 -0.327642   \n",
       "3  0.377436 -1.387024  ... -0.108300  0.005274 -0.190321 -1.175575  0.647376   \n",
       "4 -0.270533  0.817739  ... -0.009431  0.798278 -0.137458  0.141267 -0.206010   \n",
       "\n",
       "        V26       V27       V28  Class  normalizedAmount  \n",
       "0 -0.189115  0.133558 -0.021053      0          0.244964  \n",
       "1  0.125895 -0.008983  0.014724      0         -0.342475  \n",
       "2 -0.139097 -0.055353 -0.059752      0          1.160686  \n",
       "3 -0.221929  0.062723  0.061458      0          0.140534  \n",
       "4  0.502292  0.219422  0.215153      0         -0.073403  \n",
       "\n",
       "[5 rows x 31 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>V9</th>\n",
       "      <th>V10</th>\n",
       "      <th>...</th>\n",
       "      <th>V21</th>\n",
       "      <th>V22</th>\n",
       "      <th>V23</th>\n",
       "      <th>V24</th>\n",
       "      <th>V25</th>\n",
       "      <th>V26</th>\n",
       "      <th>V27</th>\n",
       "      <th>V28</th>\n",
       "      <th>Class</th>\n",
       "      <th>normalizedAmount</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.359807</td>\n",
       "      <td>-0.072781</td>\n",
       "      <td>2.536347</td>\n",
       "      <td>1.378155</td>\n",
       "      <td>-0.338321</td>\n",
       "      <td>0.462388</td>\n",
       "      <td>0.239599</td>\n",
       "      <td>0.098698</td>\n",
       "      <td>0.363787</td>\n",
       "      <td>0.090794</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.018307</td>\n",
       "      <td>0.277838</td>\n",
       "      <td>-0.110474</td>\n",
       "      <td>0.066928</td>\n",
       "      <td>0.128539</td>\n",
       "      <td>-0.189115</td>\n",
       "      <td>0.133558</td>\n",
       "      <td>-0.021053</td>\n",
       "      <td>0</td>\n",
       "      <td>0.244964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.191857</td>\n",
       "      <td>0.266151</td>\n",
       "      <td>0.166480</td>\n",
       "      <td>0.448154</td>\n",
       "      <td>0.060018</td>\n",
       "      <td>-0.082361</td>\n",
       "      <td>-0.078803</td>\n",
       "      <td>0.085102</td>\n",
       "      <td>-0.255425</td>\n",
       "      <td>-0.166974</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.225775</td>\n",
       "      <td>-0.638672</td>\n",
       "      <td>0.101288</td>\n",
       "      <td>-0.339846</td>\n",
       "      <td>0.167170</td>\n",
       "      <td>0.125895</td>\n",
       "      <td>-0.008983</td>\n",
       "      <td>0.014724</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.342475</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.358354</td>\n",
       "      <td>-1.340163</td>\n",
       "      <td>1.773209</td>\n",
       "      <td>0.379780</td>\n",
       "      <td>-0.503198</td>\n",
       "      <td>1.800499</td>\n",
       "      <td>0.791461</td>\n",
       "      <td>0.247676</td>\n",
       "      <td>-1.514654</td>\n",
       "      <td>0.207643</td>\n",
       "      <td>...</td>\n",
       "      <td>0.247998</td>\n",
       "      <td>0.771679</td>\n",
       "      <td>0.909412</td>\n",
       "      <td>-0.689281</td>\n",
       "      <td>-0.327642</td>\n",
       "      <td>-0.139097</td>\n",
       "      <td>-0.055353</td>\n",
       "      <td>-0.059752</td>\n",
       "      <td>0</td>\n",
       "      <td>1.160686</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.966272</td>\n",
       "      <td>-0.185226</td>\n",
       "      <td>1.792993</td>\n",
       "      <td>-0.863291</td>\n",
       "      <td>-0.010309</td>\n",
       "      <td>1.247203</td>\n",
       "      <td>0.237609</td>\n",
       "      <td>0.377436</td>\n",
       "      <td>-1.387024</td>\n",
       "      <td>-0.054952</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.108300</td>\n",
       "      <td>0.005274</td>\n",
       "      <td>-0.190321</td>\n",
       "      <td>-1.175575</td>\n",
       "      <td>0.647376</td>\n",
       "      <td>-0.221929</td>\n",
       "      <td>0.062723</td>\n",
       "      <td>0.061458</td>\n",
       "      <td>0</td>\n",
       "      <td>0.140534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.158233</td>\n",
       "      <td>0.877737</td>\n",
       "      <td>1.548718</td>\n",
       "      <td>0.403034</td>\n",
       "      <td>-0.407193</td>\n",
       "      <td>0.095921</td>\n",
       "      <td>0.592941</td>\n",
       "      <td>-0.270533</td>\n",
       "      <td>0.817739</td>\n",
       "      <td>0.753074</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.009431</td>\n",
       "      <td>0.798278</td>\n",
       "      <td>-0.137458</td>\n",
       "      <td>0.141267</td>\n",
       "      <td>-0.206010</td>\n",
       "      <td>0.502292</td>\n",
       "      <td>0.219422</td>\n",
       "      <td>0.215153</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.073403</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         V1        V2        V3        V4        V5        V6        V7  \\\n",
       "0 -1.359807 -0.072781  2.536347  1.378155 -0.338321  0.462388  0.239599   \n",
       "1  1.191857  0.266151  0.166480  0.448154  0.060018 -0.082361 -0.078803   \n",
       "2 -1.358354 -1.340163  1.773209  0.379780 -0.503198  1.800499  0.791461   \n",
       "3 -0.966272 -0.185226  1.792993 -0.863291 -0.010309  1.247203  0.237609   \n",
       "4 -1.158233  0.877737  1.548718  0.403034 -0.407193  0.095921  0.592941   \n",
       "\n",
       "         V8        V9       V10  ...       V21       V22       V23       V24  \\\n",
       "0  0.098698  0.363787  0.090794  ... -0.018307  0.277838 -0.110474  0.066928   \n",
       "1  0.085102 -0.255425 -0.166974  ... -0.225775 -0.638672  0.101288 -0.339846   \n",
       "2  0.247676 -1.514654  0.207643  ...  0.247998  0.771679  0.909412 -0.689281   \n",
       "3  0.377436 -1.387024 -0.054952  ... -0.108300  0.005274 -0.190321 -1.175575   \n",
       "4 -0.270533  0.817739  0.753074  ... -0.009431  0.798278 -0.137458  0.141267   \n",
       "\n",
       "        V25       V26       V27       V28  Class  normalizedAmount  \n",
       "0  0.128539 -0.189115  0.133558 -0.021053      0          0.244964  \n",
       "1  0.167170  0.125895 -0.008983  0.014724      0         -0.342475  \n",
       "2 -0.327642 -0.139097 -0.055353 -0.059752      0          1.160686  \n",
       "3  0.647376 -0.221929  0.062723  0.061458      0          0.140534  \n",
       "4 -0.206010  0.502292  0.219422  0.215153      0         -0.073403  \n",
       "\n",
       "[5 rows x 30 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = data.drop(['Time'],axis=1)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data.iloc[:, data.columns != 'Class']\n",
    "y = data.iloc[:, data.columns == 'Class']  # Response variable determining if fraudulent or not"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>Class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Class\n",
       "0      0\n",
       "1      0\n",
       "2      0\n",
       "3      0\n",
       "4      0"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X,y, test_size = 0.3, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(199364, 29)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(85443, 29)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "random_forest = RandomForestClassifier(n_estimators=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier()"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Pandas Series.ravel() function returns the flattened underlying data as an ndarray.\n",
    "random_forest.fit(X_train,y_train.values.ravel())    # np.ravel() Return a contiguous flattened array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = random_forest.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9995435553526912"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_forest.score(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Approach to plot confusion matrix (from scikit-learn.org site)\n",
    "import matplotlib.pyplot as plt\n",
    "import itertools\n",
    "\n",
    "from sklearn import svm, datasets\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "def plot_confusion_matrix(cm, classes,\n",
    "                          normalize=False,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "\n",
    "    print(cm)\n",
    "\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], fmt),\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')\n",
    "    plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[85290     6]\n",
      " [   33   114]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Confusion matrix on the test dataset\n",
    "cnf_matrix = confusion_matrix(y_test,y_pred)\n",
    "plot_confusion_matrix(cnf_matrix,classes=[0,1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- while only 6 regular transactions are wrongly predicted as fraudulent, the model only detects 78% of the fraudulent transactions. As a consequence 33 fraudulent transactions are not detected (False Negatives).\n",
    "- Let's see if we can improve this performance with other machine learning / deep learning models in the rest of the notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy:0.9995 \tprecision:0.9500 \trecall:0.7755 \tF1-score:0.8539\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score, f1_score, precision_score, recall_score, plot_roc_curve\n",
    "acc = accuracy_score(y_test, y_pred)\n",
    "prec = precision_score(y_test, y_pred)\n",
    "rec = recall_score(y_test, y_pred)\n",
    "f1 = f1_score(y_test, y_pred)\n",
    "print('accuracy:%0.4f'%acc,'\\tprecision:%0.4f'%prec,'\\trecall:%0.4f'%rec,'\\tF1-score:%0.4f'%f1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's store each model's performance in a dataframe for comparison purpose"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999544</td>\n",
       "      <td>0.22449</td>\n",
       "      <td>0.77551</td>\n",
       "      <td>0.95</td>\n",
       "      <td>0.853933</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Model  Accuracy  FalseNegRate   Recall  Precision  F1 Score\n",
       "0  RandomForest  0.999544       0.22449  0.77551       0.95  0.853933"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Store results in dataframe for comparing various Models\n",
    "results_testset = pd.DataFrame([['RandomForest', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_testset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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": [
    "ROC_RF = plot_roc_curve(random_forest, X_test, y_test)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will run the models on the full dataset to check."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[284309      6]\n",
      " [    34    458]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Confusion matrix on the whole dataset\n",
    "y_pred = random_forest.predict(X)\n",
    "cnf_matrix = confusion_matrix(y,y_pred.round())\n",
    "plot_confusion_matrix(cnf_matrix,classes=[0,1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy:0.9999 \tprecision:0.9871 \trecall:0.9309 \tF1-score:0.9582\n"
     ]
    }
   ],
   "source": [
    "acc = accuracy_score(y, y_pred)\n",
    "prec = precision_score(y, y_pred)\n",
    "rec = recall_score(y, y_pred)\n",
    "f1 = f1_score(y, y_pred)\n",
    "print('accuracy:%0.4f'%acc,'\\tprecision:%0.4f'%prec,'\\trecall:%0.4f'%rec,'\\tF1-score:%0.4f'%f1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.99986</td>\n",
       "      <td>0.069106</td>\n",
       "      <td>0.930894</td>\n",
       "      <td>0.987069</td>\n",
       "      <td>0.958159</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Model  Accuracy  FalseNegRate    Recall  Precision  F1 Score\n",
       "0  RandomForest   0.99986      0.069106  0.930894   0.987069  0.958159"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_fullset = pd.DataFrame([['RandomForest', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_fullset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Decision trees"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "decision_tree = DecisionTreeClassifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier()"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "decision_tree.fit(X_train,y_train.values.ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = decision_tree.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9992392589211521"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "decision_tree.score(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[85267    29]\n",
      " [   36   111]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Confusion matrix on the test dataset\n",
    "cnf_matrix = confusion_matrix(y_test,y_pred)\n",
    "plot_confusion_matrix(cnf_matrix,classes=[0,1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- The performance of the Decision Tree model is below the one using Random Forest. Let's check the performance indicators."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = accuracy_score(y_test, y_pred)\n",
    "prec = precision_score(y_test, y_pred)\n",
    "rec = recall_score(y_test, y_pred)\n",
    "f1 = f1_score(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999544</td>\n",
       "      <td>0.224490</td>\n",
       "      <td>0.775510</td>\n",
       "      <td>0.950000</td>\n",
       "      <td>0.853933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.999239</td>\n",
       "      <td>0.244898</td>\n",
       "      <td>0.755102</td>\n",
       "      <td>0.792857</td>\n",
       "      <td>0.773519</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Model  Accuracy  FalseNegRate    Recall  Precision  F1 Score\n",
       "0  RandomForest  0.999544      0.224490  0.775510   0.950000  0.853933\n",
       "1  DecisionTree  0.999239      0.244898  0.755102   0.792857  0.773519"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Store results in dataframe for comparing various Models\n",
    "model_results = pd.DataFrame([['DecisionTree', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_testset = results_testset.append(model_results, ignore_index = True)\n",
    "results_testset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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": [
    "ROC_DT = plot_roc_curve(decision_tree, X_test, y_test)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[284286     29]\n",
      " [    36    456]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Confusion matrix on the whole dataset\n",
    "y_pred = decision_tree.predict(X)\n",
    "cnf_matrix = confusion_matrix(y,y_pred.round())\n",
    "plot_confusion_matrix(cnf_matrix,classes=[0,1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = accuracy_score(y, y_pred)\n",
    "prec = precision_score(y, y_pred)\n",
    "rec = recall_score(y, y_pred)\n",
    "f1 = f1_score(y, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999860</td>\n",
       "      <td>0.069106</td>\n",
       "      <td>0.930894</td>\n",
       "      <td>0.987069</td>\n",
       "      <td>0.958159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.999772</td>\n",
       "      <td>0.073171</td>\n",
       "      <td>0.926829</td>\n",
       "      <td>0.940206</td>\n",
       "      <td>0.933470</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Model  Accuracy  FalseNegRate    Recall  Precision  F1 Score\n",
       "0  RandomForest  0.999860      0.069106  0.930894   0.987069  0.958159\n",
       "1  DecisionTree  0.999772      0.073171  0.926829   0.940206  0.933470"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_results = pd.DataFrame([['DecisionTree', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_fullset = results_fullset.append(model_results, ignore_index = True)\n",
    "results_fullset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Let's now explore Neural Network models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = np.array(X_train)\n",
    "X_test = np.array(X_test)\n",
    "y_train = np.array(y_train)\n",
    "y_test = np.array(y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Deep neural network\n",
    "- We will use a simple NN made of 5 fully-connected layers with ReLu activation. The NN takes a vector of length 29 as input. This represents the information related to each transactions, ie each line with 29 columns from the dataset. For each transaction, the final layer will output a probability distribution (sigmoid activation function) and classify either as not fraudulent (0) or fraudulent (1).\n",
    "- a dropout step is included to prevent overfitting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential([\n",
    "    Dense(units=16, input_dim = 29,activation='relu'),   # input of 29 columns as shown above\n",
    "    Dense(units=24,activation='relu'),\n",
    "    Dropout(0.5),\n",
    "    Dense(24,activation='relu'),\n",
    "    Dense(24,activation='relu'),\n",
    "    Dense(1,activation='sigmoid'),                        # binary classification fraudulent or not\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_1 (Dense)              (None, 16)                480       \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 24)                408       \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 24)                0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 24)                600       \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 24)                600       \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 1)                 25        \n",
      "=================================================================\n",
      "Total params: 2,113\n",
      "Trainable params: 2,113\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "199364/199364 [==============================] - 35s 175us/step - loss: 0.0088 - accuracy: 0.9983\n",
      "Epoch 2/5\n",
      "199364/199364 [==============================] - 35s 174us/step - loss: 0.0041 - accuracy: 0.9994\n",
      "Epoch 3/5\n",
      "199364/199364 [==============================] - 34s 172us/step - loss: 0.0036 - accuracy: 0.9994\n",
      "Epoch 4/5\n",
      "199364/199364 [==============================] - 35s 173us/step - loss: 0.0033 - accuracy: 0.9994\n",
      "Epoch 5/5\n",
      "199364/199364 [==============================] - 35s 176us/step - loss: 0.0034 - accuracy: 0.9994\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x19d45fa8a08>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.compile(optimizer='adam',loss='binary_crossentropy',metrics=['accuracy'])\n",
    "model.fit(X_train,y_train,batch_size=15,epochs=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "85443/85443 [==============================] - 3s 37us/step\n"
     ]
    }
   ],
   "source": [
    "score = model.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.0039969228647745, 0.9993796944618225]\n"
     ]
    }
   ],
   "source": [
    "print(score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- The model achieves an accuracy of 99.94% ! Is this a good performance ?\n",
    "- Remember that our dataset is significantly composed of non fraudulent samples with only 172 fraudulent transactions per 100,000. Consequently, a model predicting every transaction as 'non fraudulent' would achieve 99.83% accuracy despite being unable to detect a single fraudulent case !"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Data Accuracy: 0.9994\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Confusion Matrix on unsee test set\n",
    "import seaborn as sn\n",
    "y_pred = model.predict(X_test)\n",
    "for i in range(len(y_test)):\n",
    "    if y_pred[i]>0.5:\n",
    "        y_pred[i]=1 \n",
    "    else:\n",
    "        y_pred[i]=0\n",
    "cm = confusion_matrix(y_test, y_pred) # rows = truth, cols = prediction\n",
    "df_cm = pd.DataFrame(cm, index = (0, 1), columns = (0, 1))\n",
    "plt.figure(figsize = (10,7))\n",
    "#sn.set(font_scale=1.4)\n",
    "sn.heatmap(df_cm, annot=True, fmt='g')\n",
    "print(\"Test Data Accuracy: %0.4f\" % accuracy_score(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Detection of fraudulent transactions did not improve compared to the previous machine learning models.\n",
    "- 115 fraudulent transactions are detected as fraudulent by the model, yet 32 fraudulent transactions are not identified (false negative) which remains an issue. Our objective must be to detect as many fraudulent transactions as possible since these can have a huge negative impact.\n",
    "- 21 regular transactions are detected as potentially fraudulent by the model. These are false positive. This number is negligible.\n",
    "\n",
    "Conclusion : We must find ways to further reduce the number of false negative."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Alternative approach to plot confusion matrix (from scikit-learn.org site)\n",
    "y_pred = model.predict(X_test)\n",
    "y_test = pd.DataFrame(y_test)    # Pandas format required by confusion_matrix function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "cnf_matrix = confusion_matrix(y_test, y_pred.round())   # y_pred.round() to convert probability to either 0 or 1 in line with y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[85275    21]\n",
      " [   32   115]]\n"
     ]
    }
   ],
   "source": [
    "print(cnf_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[85275    21]\n",
      " [   32   115]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(cnf_matrix, classes=[0,1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = accuracy_score(y_test, y_pred.round())\n",
    "prec = precision_score(y_test, y_pred.round())\n",
    "rec = recall_score(y_test, y_pred.round())\n",
    "f1 = f1_score(y_test, y_pred.round())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999544</td>\n",
       "      <td>0.224490</td>\n",
       "      <td>0.775510</td>\n",
       "      <td>0.950000</td>\n",
       "      <td>0.853933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.999239</td>\n",
       "      <td>0.244898</td>\n",
       "      <td>0.755102</td>\n",
       "      <td>0.792857</td>\n",
       "      <td>0.773519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PlainNeuralNetwork</td>\n",
       "      <td>0.999380</td>\n",
       "      <td>0.217687</td>\n",
       "      <td>0.782313</td>\n",
       "      <td>0.845588</td>\n",
       "      <td>0.812721</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Model  Accuracy  FalseNegRate    Recall  Precision  F1 Score\n",
       "0        RandomForest  0.999544      0.224490  0.775510   0.950000  0.853933\n",
       "1        DecisionTree  0.999239      0.244898  0.755102   0.792857  0.773519\n",
       "2  PlainNeuralNetwork  0.999380      0.217687  0.782313   0.845588  0.812721"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Store results in dataframe for comparing various Models\n",
    "model_results = pd.DataFrame([['PlainNeuralNetwork', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_testset = results_testset.append(model_results, ignore_index = True)\n",
    "results_testset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[284249     66]\n",
      " [   104    388]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Confusion matrix on the whole dataset\n",
    "y_pred = model.predict(X)\n",
    "y_expected = pd.DataFrame(y)\n",
    "cnf_matrix = confusion_matrix(y_expected, y_pred.round())\n",
    "plot_confusion_matrix(cnf_matrix,classes=[0,1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = accuracy_score(y, y_pred.round())\n",
    "prec = precision_score(y, y_pred.round())\n",
    "rec = recall_score(y, y_pred.round())\n",
    "f1 = f1_score(y, y_pred.round())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999860</td>\n",
       "      <td>0.069106</td>\n",
       "      <td>0.930894</td>\n",
       "      <td>0.987069</td>\n",
       "      <td>0.958159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.999772</td>\n",
       "      <td>0.073171</td>\n",
       "      <td>0.926829</td>\n",
       "      <td>0.940206</td>\n",
       "      <td>0.933470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PlainNeuralNetwork</td>\n",
       "      <td>0.999403</td>\n",
       "      <td>0.211382</td>\n",
       "      <td>0.788618</td>\n",
       "      <td>0.854626</td>\n",
       "      <td>0.820296</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Model  Accuracy  FalseNegRate    Recall  Precision  F1 Score\n",
       "0        RandomForest  0.999860      0.069106  0.930894   0.987069  0.958159\n",
       "1        DecisionTree  0.999772      0.073171  0.926829   0.940206  0.933470\n",
       "2  PlainNeuralNetwork  0.999403      0.211382  0.788618   0.854626  0.820296"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_results = pd.DataFrame([['PlainNeuralNetwork', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_fullset = results_fullset.append(model_results, ignore_index = True)\n",
    "results_fullset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Weighted loss to account for large class imbalance in train dataset\n",
    "- we will adjust the class imbalance by giving additional weight to the loss associated to errors made on fraudulent transaction detection. Let's review the process:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lveys\\anaconda3\\envs\\tensorflow_py37\\lib\\site-packages\\sklearn\\utils\\validation.py:70: FutureWarning: Pass classes=[0 1], y=[0 0 0 ... 0 0 0] as keyword args. From version 0.25 passing these as positional arguments will result in an error\n",
      "  FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{0: 0.500866751415694, 1: 288.93333333333334}"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.utils import class_weight\n",
    "class_weights = class_weight.compute_class_weight('balanced', np.unique(y_train), np.array([y_train[i][0] for i in range(len(y_train))]))\n",
    "class_weights = dict(enumerate(class_weights))\n",
    "class_weights"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- The class 'Fraudulent' (y=1) is assigned a weight of 289 vs 0.5 for the class 'not fraudulent' due to the very low prevalence we detected during data exploration. This allows the model to give more importance to the errors made on fraudulent cases during training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "199364/199364 [==============================] - 35s 174us/step - loss: 0.4578 - accuracy: 0.9874\n",
      "Epoch 2/5\n",
      "199364/199364 [==============================] - 35s 177us/step - loss: 0.4431 - accuracy: 0.9728\n",
      "Epoch 3/5\n",
      "199364/199364 [==============================] - 36s 179us/step - loss: 0.2458 - accuracy: 0.9690\n",
      "Epoch 4/5\n",
      "199364/199364 [==============================] - 35s 178us/step - loss: 0.2681 - accuracy: 0.9723\n",
      "Epoch 5/5\n",
      "199364/199364 [==============================] - 35s 175us/step - loss: 0.2549 - accuracy: 0.9726\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x19d44432a48>"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train,y_train,batch_size=15,epochs=5, class_weight=class_weights, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "85443/85443 [==============================] - 3s 37us/step\n"
     ]
    }
   ],
   "source": [
    "score_weighted = model.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.05213862410299367, 0.9759371876716614]\n"
     ]
    }
   ],
   "source": [
    "print(score_weighted)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Data Accuracy: 0.9759\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Confusion Matrix on unseen test set\n",
    "y_pred = model.predict(X_test)\n",
    "for i in range(len(y_test)):\n",
    "    if y_pred[i]>0.5:\n",
    "        y_pred[i]=1 \n",
    "    else:\n",
    "        y_pred[i]=0\n",
    "cm = confusion_matrix(y_test, y_pred) # rows = truth, cols = prediction\n",
    "df_cm = pd.DataFrame(cm, index = (0, 1), columns = (0, 1))\n",
    "plt.figure(figsize = (10,7))\n",
    "#sn.set(font_scale=1.4)\n",
    "sn.heatmap(df_cm, annot=True, fmt='g')\n",
    "print(\"Test Data Accuracy: %0.4f\" % accuracy_score(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- The accuracy reduced a bit. We need to look at more detailed evaluation measures like precision and recall to gauge the true model performance.\n",
    "- The model is actually much better at detecting fraudulent cases now. We have a lower False negative rate which is the key criteria for our purpose (detect a fraud when there is one).\n",
    "- But on the other hand, the model generates an excessive number of false positive compared to the previous approaches. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = accuracy_score(y_test, y_pred.round())\n",
    "prec = precision_score(y_test, y_pred.round())\n",
    "rec = recall_score(y_test, y_pred.round())\n",
    "f1 = f1_score(y_test, y_pred.round())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999544</td>\n",
       "      <td>0.224490</td>\n",
       "      <td>0.775510</td>\n",
       "      <td>0.950000</td>\n",
       "      <td>0.853933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.999239</td>\n",
       "      <td>0.244898</td>\n",
       "      <td>0.755102</td>\n",
       "      <td>0.792857</td>\n",
       "      <td>0.773519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PlainNeuralNetwork</td>\n",
       "      <td>0.999380</td>\n",
       "      <td>0.217687</td>\n",
       "      <td>0.782313</td>\n",
       "      <td>0.845588</td>\n",
       "      <td>0.812721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>WeightedNeuralNetwork</td>\n",
       "      <td>0.975937</td>\n",
       "      <td>0.102041</td>\n",
       "      <td>0.897959</td>\n",
       "      <td>0.060746</td>\n",
       "      <td>0.113793</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Model  Accuracy  FalseNegRate    Recall  Precision  \\\n",
       "0           RandomForest  0.999544      0.224490  0.775510   0.950000   \n",
       "1           DecisionTree  0.999239      0.244898  0.755102   0.792857   \n",
       "2     PlainNeuralNetwork  0.999380      0.217687  0.782313   0.845588   \n",
       "3  WeightedNeuralNetwork  0.975937      0.102041  0.897959   0.060746   \n",
       "\n",
       "   F1 Score  \n",
       "0  0.853933  \n",
       "1  0.773519  \n",
       "2  0.812721  \n",
       "3  0.113793  "
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Store results in dataframe for comparing various Models\n",
    "model_results = pd.DataFrame([['WeightedNeuralNetwork', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_testset = results_testset.append(model_results, ignore_index = True)\n",
    "results_testset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Although the recall metric improves, the F1-score collapses due to extremely poor precision. Given the very high number of transactions processed, the excessive number of false positives is clearly an issue."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[277603   6712]\n",
      " [    24    468]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Confusion matrix on the whole dataset\n",
    "y_pred = model.predict(X)\n",
    "y_expected = pd.DataFrame(y)\n",
    "cnf_matrix = confusion_matrix(y_expected, y_pred.round())\n",
    "plot_confusion_matrix(cnf_matrix,classes=[0,1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = accuracy_score(y, y_pred.round())\n",
    "prec = precision_score(y, y_pred.round())\n",
    "rec = recall_score(y, y_pred.round())\n",
    "f1 = f1_score(y, y_pred.round())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999860</td>\n",
       "      <td>0.069106</td>\n",
       "      <td>0.930894</td>\n",
       "      <td>0.987069</td>\n",
       "      <td>0.958159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.999772</td>\n",
       "      <td>0.073171</td>\n",
       "      <td>0.926829</td>\n",
       "      <td>0.940206</td>\n",
       "      <td>0.933470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PlainNeuralNetwork</td>\n",
       "      <td>0.999403</td>\n",
       "      <td>0.211382</td>\n",
       "      <td>0.788618</td>\n",
       "      <td>0.854626</td>\n",
       "      <td>0.820296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>WeightedNeuralNetwork</td>\n",
       "      <td>0.976349</td>\n",
       "      <td>0.048780</td>\n",
       "      <td>0.951220</td>\n",
       "      <td>0.065181</td>\n",
       "      <td>0.122002</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Model  Accuracy  FalseNegRate    Recall  Precision  \\\n",
       "0           RandomForest  0.999860      0.069106  0.930894   0.987069   \n",
       "1           DecisionTree  0.999772      0.073171  0.926829   0.940206   \n",
       "2     PlainNeuralNetwork  0.999403      0.211382  0.788618   0.854626   \n",
       "3  WeightedNeuralNetwork  0.976349      0.048780  0.951220   0.065181   \n",
       "\n",
       "   F1 Score  \n",
       "0  0.958159  \n",
       "1  0.933470  \n",
       "2  0.820296  \n",
       "3  0.122002  "
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_results = pd.DataFrame([['WeightedNeuralNetwork', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_fullset = results_fullset.append(model_results, ignore_index = True)\n",
    "results_fullset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Undersampling\n",
    "- In order to balance the train set, another technique is undersampling. With this technique, we adjust the largest class to match the number of samples of the under-represented class. Here we want to randomly pick an amount of non-fraudulent transactions equal to the number of fraudulent transactions in the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "492\n"
     ]
    }
   ],
   "source": [
    "fraud_indices = np.array(data[data.Class == 1].index)\n",
    "number_records_fraud = len(fraud_indices)\n",
    "print(number_records_fraud)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "normal_indices = data[data.Class == 0].index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "284315"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(normal_indices)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "492\n"
     ]
    }
   ],
   "source": [
    "# Random select N indices from non fraudulent samples (N equals to number of fraudulent records)\n",
    "random_normal_indices = np.random.choice(normal_indices, number_records_fraud, replace=False)\n",
    "random_normal_indices = np.array(random_normal_indices)\n",
    "print(len(random_normal_indices))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "984\n"
     ]
    }
   ],
   "source": [
    "under_sample_indices = np.concatenate([fraud_indices,random_normal_indices])\n",
    "print(len(under_sample_indices))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "under_sample_data = data.iloc[under_sample_indices,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_undersample = under_sample_data.iloc[:,under_sample_data.columns != 'Class']\n",
    "y_undersample = under_sample_data.iloc[:,under_sample_data.columns == 'Class']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X_undersample,y_undersample, test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = np.array(X_train)\n",
    "X_test = np.array(X_test)\n",
    "y_train = np.array(y_train)\n",
    "y_test = np.array(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_1 (Dense)              (None, 16)                480       \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 24)                408       \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 24)                0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 24)                600       \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 24)                600       \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 1)                 25        \n",
      "=================================================================\n",
      "Total params: 2,113\n",
      "Trainable params: 2,113\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "688/688 [==============================] - 0s 408us/step - loss: 0.2261 - accuracy: 0.9491\n",
      "Epoch 2/5\n",
      "688/688 [==============================] - 0s 198us/step - loss: 0.2028 - accuracy: 0.9506\n",
      "Epoch 3/5\n",
      "688/688 [==============================] - 0s 190us/step - loss: 0.1509 - accuracy: 0.9477\n",
      "Epoch 4/5\n",
      "688/688 [==============================] - 0s 190us/step - loss: 0.1238 - accuracy: 0.9608\n",
      "Epoch 5/5\n",
      "688/688 [==============================] - 0s 187us/step - loss: 0.1096 - accuracy: 0.9666\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x19d270bdb88>"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.compile(optimizer='adam',loss='binary_crossentropy',metrics=['accuracy'])\n",
    "model.fit(X_train,y_train,batch_size=15,epochs=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[141   6]\n",
      " [  6 143]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = model.predict(X_test)\n",
    "y_expected = pd.DataFrame(y_test)\n",
    "cnf_matrix = confusion_matrix(y_expected, y_pred.round())\n",
    "plot_confusion_matrix(cnf_matrix, classes=[0,1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- The metrics look pretty strong on the test set...."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = accuracy_score(y_test, y_pred.round())\n",
    "prec = precision_score(y_test, y_pred.round())\n",
    "rec = recall_score(y_test, y_pred.round())\n",
    "f1 = f1_score(y_test, y_pred.round())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999544</td>\n",
       "      <td>0.224490</td>\n",
       "      <td>0.775510</td>\n",
       "      <td>0.950000</td>\n",
       "      <td>0.853933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.999239</td>\n",
       "      <td>0.244898</td>\n",
       "      <td>0.755102</td>\n",
       "      <td>0.792857</td>\n",
       "      <td>0.773519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PlainNeuralNetwork</td>\n",
       "      <td>0.999380</td>\n",
       "      <td>0.217687</td>\n",
       "      <td>0.782313</td>\n",
       "      <td>0.845588</td>\n",
       "      <td>0.812721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>WeightedNeuralNetwork</td>\n",
       "      <td>0.975937</td>\n",
       "      <td>0.102041</td>\n",
       "      <td>0.897959</td>\n",
       "      <td>0.060746</td>\n",
       "      <td>0.113793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>UnderSampledNeuralNetwork</td>\n",
       "      <td>0.959459</td>\n",
       "      <td>0.040268</td>\n",
       "      <td>0.959732</td>\n",
       "      <td>0.959732</td>\n",
       "      <td>0.959732</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Model  Accuracy  FalseNegRate    Recall  Precision  \\\n",
       "0               RandomForest  0.999544      0.224490  0.775510   0.950000   \n",
       "1               DecisionTree  0.999239      0.244898  0.755102   0.792857   \n",
       "2         PlainNeuralNetwork  0.999380      0.217687  0.782313   0.845588   \n",
       "3      WeightedNeuralNetwork  0.975937      0.102041  0.897959   0.060746   \n",
       "4  UnderSampledNeuralNetwork  0.959459      0.040268  0.959732   0.959732   \n",
       "\n",
       "   F1 Score  \n",
       "0  0.853933  \n",
       "1  0.773519  \n",
       "2  0.812721  \n",
       "3  0.113793  \n",
       "4  0.959732  "
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Store results in dataframe for comparing various Models\n",
    "model_results = pd.DataFrame([['UnderSampledNeuralNetwork', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_testset = results_testset.append(model_results, ignore_index = True)\n",
    "results_testset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "....But the model fails on non fraudulent transactions as shown below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[274591   9724]\n",
      " [    16    476]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU8AAAEYCAYAAADcRnS9AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAlZUlEQVR4nO3debxVVf3/8debMVBBBjEETVPEKUFB0EwzLSGywJzIiRR/DmnlN620/Krpl+ayzClNFGfIEUck0BRnRHJARUzSqySjiAPoxc/vj70uHi53OMMduPe8nz32w3PW3mvtz4H8uPZae6+tiMDMzArTprkDMDNriZw8zcyK4ORpZlYEJ08zsyI4eZqZFcHJ08ysCE6erYykTpLulLRc0t9LaOcISfc3ZGzNRdJekl5u7jisdZHv82wekg4HfgRsB6wAZgPjImJGie0eBXwf+GJEVJYa5/pOUgD9ImJec8di5cU9z2Yg6UfAn4BfApsCWwCXACMboPnPAXPLIXHmQ1K75o7BWqmI8NaEG9AVeA84pI5jOpIl17fS9iegY9q3D1ABnAYsBBYAx6R9vwA+Aj5O5xgLnAtcl9P2lkAA7dL37wL/Juv9vgYckVM+I6feF4GngOXpn1/M2fcgcD7wSGrnfqBnLb+tKv6f5MQ/ChgBzAWWAj/LOX4I8BjwTjr2IqBD2vdQ+i3vp997WE77PwX+C1xbVZbqbJ3OsWv6vhmwGNinuf+/4a1lbe55Nr09gM8At9VxzM+B3YGBwACyBHJWzv7PkiXhPmQJ8mJJ3SLiHLLe7MSI2DAirqwrEEkbABcCX4+IjcgS5OwajusO3J2O7QH8EbhbUo+cww4HjgF6AR2A0+s49WfJ/gz6AGcDVwBHAoOAvYCzJX0+Hbsa+B+gJ9mf3X7A9wAiYu90zID0eyfmtN+drBd+fO6JI+JVssR6vaTOwFXA1RHxYB3xmq3DybPp9QAWR92X1UcA50XEwohYRNajPCpn/8dp/8cRcQ9Zr6t/kfF8AuwkqVNELIiIF2o45hvAKxFxbURURsSNwEvAN3OOuSoi5kbEh8AkssRfm4/Jxnc/Bm4iS4x/jogV6fwvADsDRMTTEfF4Ou984K/Al/P4TedExKoUz1oi4grgFeAJoDfZf6zMCuLk2fSWAD3rGYvbDPhPzvf/pLI1bVRLvh8AGxYaSES8T3apeyKwQNLdkrbLI56qmPrkfP9vAfEsiYjV6XNVcns7Z/+HVfUlbSvpLkn/lfQuWc+6Zx1tAyyKiJX1HHMFsBPwl4hYVc+xZutw8mx6jwErycb5avMW2SVnlS1SWTHeBzrnfP9s7s6ImBIRXyPrgb1EllTqi6cqpjeLjKkQl5LF1S8iugA/A1RPnTpvIZG0Idk48pXAuWlYwqwgTp5NLCKWk43zXSxplKTOktpL+rqk36bDbgTOkrSJpJ7p+OuKPOVsYG9JW0jqCpxZtUPSppK+lcY+V5Fd/q+uoY17gG0lHS6pnaTDgB2Au4qMqRAbAe8C76Ve8UnV9r8NfH6dWnX7M/B0RBxHNpZ7WclRWtlx8mwGEfFHsns8zwIWAW8ApwC3p0P+D5gJPAs8B8xKZcWcayowMbX1NGsnvDZks/Zvkc1Af5k0GVOtjSXAAenYJWQz5QdExOJiYirQ6WSTUSvIesUTq+0/F5gg6R1Jh9bXmKSRwHCyoQrI/h52lXREg0VsZcE3yZuZFcE9TzOzIjh5mpkVwcnTzKwITp5mZkVYrxZNULtOoQ4bNXcYVoAB223R3CFYAV5/fT5LFi+u7z7ZgrTt8rmIynUe5KpRfLhoSkQMb8jzN5f1K3l22IiO/eu928TWIw888ufmDsEK8JU9hzZ4m1H5Yd7/3q6cfXF9T4e1GOtV8jSzlkig8hsBdPI0s9IIaNO2uaNock6eZlY6Negwaovg5GlmJfJlu5lZcdzzNDMrkHDP08yscHLP08ysKJ5tNzMrlCeMzMwKJ3zZbmZWFPc8zcwK5ct2M7PitPFlu5lZYfxsu5lZMcrzsr38frGZNTwpv63OJrS5pAckvSjpBUk/TOXnSnpT0uy0jcipc6akeZJeljQsp3yQpOfSvgul7OSSOkqamMqfkLRlTp0xkl5J25j6frJ7nmZWuobpeVYCp0XELEkbAU9Lmpr2XRARv1/rlNIOwGhgR2Az4B+Sto2I1cClwPHA48A9wHDgXmAssCwitpE0GvgNcJik7sA5wGAg0rknR8Sy2oJ1z9PMSpNvr7OenmdELIiIWenzCuBFoE8dVUYCN0XEqoh4DZgHDJHUG+gSEY9FRADXAKNy6kxIn28G9ku90mHA1IhYmhLmVLKEWysnTzMrXZu2+W3QU9LMnO34mppLl9O7AE+kolMkPStpvKRuqawP8EZOtYpU1id9rl6+Vp2IqASWAz3qaKv2n1zXTjOz+qUJo3w2WBwRg3O2y9dpTdoQuAU4NSLeJbsE3xoYCCwA/vDpidcRdZQXW6dGTp5mVroGuGzPmlF7ssR5fUTcChARb0fE6oj4BLgCGJIOrwA2z6neF3grlfetoXytOpLaAV2BpXW0VSsnTzMrTdV6nvn1PGtvJht7vBJ4MSL+mFPeO+ewA4Hn0+fJwOg0g74V0A94MiIWACsk7Z7aPBq4I6dO1Uz6wcD0NC46BdhfUrc0LLB/KquVZ9vNrEQNdp/nnsBRwHOSZqeynwHfkTSQ7DJ6PnACQES8IGkSMIdspv7kNNMOcBJwNdCJbJb93lR+JXCtpHlkPc7Rqa2lks4HnkrHnRcRS+sK1snTzErXAKsqRcQMah57vKeOOuOAcTWUzwR2qqF8JXBILW2NB8bnG6+Tp5mVzo9nmpkVSOX5eKaTp5mVzoshm5kVTk6eZmaFyd7C4eRpZlYYUfMceSvn5GlmJRJt2njCyMysYL5sNzMrgpOnmVmhPOZpZlY4Ifc8zcyK4QkjM7MiuOdpZlYoj3mamRXHPU8zswJ5wsjMrEhOnmZmhRKojZOnmVnB3PM0MyuCk6eZWYE8YWRmVqzyy51OnnXpu+nG/O38o9m0Rxc+iWD8LY9w8Y0Pcu2vj6HflpsCsPFGnXhnxYfsPvrXa+pt/tluzLrlLMZddg9/unYaAFOu+CGf7dmFD1d9DMA3T7qIRcveY4ve3bjsnCPp2W1Dlr37Acf+fAJvLnwHgDsu+h5Ddt6SR5/5Nwf98LKm/fGt0GUXX8iEq66ECI4+ZiwnnfJDjj3qO7wydy4Ay5e/Q9euG/PwE0/zwLSp/OJ/f85HH39Eh/YdOO+Xv2bvffZdq73vHDyK+fNf47GZ/2qOn7P+kC/brZrK1Z9wxh9vZfZLFWzYuSOP3vBTpj3xEkedcdWaY379owNZ/t6Ha9X77ekHcf8jL6zT3jE/n8CsOa+vVfar/zmQ6+9+kuvvfIIv77Yt533/W4z932sAuOCaf9D5Mx0Ye9CXGuHXlZc5LzzPhKuuZNpDj9GhQwcOHjmC/YePYPy1N6455qwzTqdLl64A9OjRkxtvvp3em23GnBee5+BvjWDOq5/+3d15+21ssOGGTf471lfl+Gx7+f3iAvx38bvMfqkCgPc+WMVLr/2XzTbZeK1jDvrarky67+k137+5z868VrGYOa/+N69zbPf53jz4xMsA/POpuRywzxfW7HvwybmseH9Vib/CAOa+/BK77TaUzp07065dO/b80t7cNfn2NfsjgttuuZmDDh0NwM4Dd6H3ZpsBsP0OO7Jy1UpWrcr+Lt577z0u/ssFnP7TnzX571hvKc+tFXHyzNMWvbszsH9fnnp+/pqyPXfdmreXruDV1xcB0PkzHTjtmK8x7q/31NjGX889ksdvOoMz/t/wNWXPzX2TUfsNBGDkvgPosmEnunfdoNF+R7nafocdefSRh1m6ZAkffPABU6fcy5sVFWv2P/rIw/TqtSlbb9NvnbqTb7+VnQcMpGPHjgD88ryzOeUHP6Jz585NFv/6TlJeW2vSqMlT0nBJL0uaJ+mMxjxXY9qgUwdu/P1x/Pj3t7Di/ZVryg8dPpi/3zdzzff/Pekb/OW66bz/4UfrtHHMz65mt0N/yVePvYA9d9maww8YAsCZF9zGXoO24bEbf8peg7bhzbeXUbl6deP/qDLTf7vt+eGPfsyBBwzn4JEj2PELA2jXru2a/bdMmshBhx62Tr0X57zAuWedyQV/uRSA5/41m3+/+ioHjBzVVKGv9/JNnK0teTbamKektsDFwNeACuApSZMjYk5jnbMxtGvXhht///+YeO9M7pj+6cRA27ZtGLnvAPY8/Ldrynbb6XMc+NWBjDt1FF036sQnnwQrP/qYyyY+xFuLlgPZ5f/Ee2ey246f44a7nmTBouWMPv1vQJakR+03kHffW4k1vKO+eyxHffdYAM47++ds1qcvAJWVldw1+TYemPHkWse/WVHBUaMP5tK/XcVWn98agCefeJx/PTOLnbfbmtWVlSxatJADhu3LXVOmN+2PWc+0tsSYj8acMBoCzIuIfwNIugkYCbSo5HnZOUfw8mv/5cLr1v6XY9+h/Zk7/+01M+MAXx37pzWff37CCN7/YBWXTXyItm3bsPFGnVjyzvu0a9eGEXvvxPQ0ztlj4w1YuvwDIoIfHzuMCXc83hQ/qywtWriQTXr14o03Xueuybdz/wMzAHhw+j/ot21/+vTtu+bY5e+8w2EHfYuzzxvH7nvsuaZ87PEnMvb4EwF4/T/zOeygkWWfOMHJs6H1Ad7I+V4BDK1+kKTjgeMBaL9+zV5+ceDnOeKAoTw3900evykbdTjnoslMmTGHQ4YNWmuiqC4d27dj8sUn075dW9q2bcMDT7zE+FsfAWDvwf047/vfIgJmzJrHqb+atKbeP648lW232pQNO3Vk3n3nc+IvbuAfj73Y8D+0TBx9+CEsW7qUdu3b87sLLmTjbt0AuPXmSRx0yOi1jr3isot57dV5/O5X4/jdr8Zlx915L5v06tXkcbcE5fhsuyKicRqWDgGGRcRx6ftRwJCI+H5tddp07hUd+x/aKPFY41jw6J+bOwQrwFf2HMozs2Y2aKbr+Nl+0feIC/M69t9/HPF0RAyuaZ+kzYFrgM8CnwCXR8SfJXUHJgJbAvOBQyNiWapzJjAWWA38ICKmpPJBwNVAJ+Ae4IcREZI6pnMMApYAh0XE/FRnDHBWCuf/ImJCXb+lMSeMKoDNc773Bd5qxPOZWTMQIOW31aMSOC0itgd2B06WtANwBjAtIvoB09J30r7RwI7AcOCSNNcCcCnZFW2/tFXd4jIWWBYR2wAXAL9JbXUHziG7Oh4CnCOpW13BNmbyfAroJ2krSR3IfuTkRjyfmTWLhpltj4gFETErfV4BvEg2/DcSqOoFTgBGpc8jgZsiYlVEvAbMA4ZI6g10iYjHIru0vqZanaq2bgb2UxbYMGBqRCxNvdqpfJpwa9RoY54RUSnpFGAK0BYYHxHrPnZjZi1eQ88XSdoS2AV4Atg0IhZAlmAlVQ089wFyZ1grUtnH6XP18qo6b6S2KiUtB3pQ8xxNH+rQqI9nRsQ9ZOMNZtZaCdrkP2HUU9LMnO+XR8TlazUnbQjcApwaEe/W0WOtaUfUUV5snRr52XYzK4koKHkurm3CCEBSe7LEeX1E3JqK35bUO/U6ewMLU3lt8yoV6XP18tw6FZLaAV2Bpal8n2p1Hqzrh/jxTDMrWUNMGKWxxyuBFyPijzm7JgNj0ucxwB055aMldZS0FdnE0JPpEn+FpN1Tm0dXq1PV1sHA9DQuOgXYX1K3NFG0fyqrlXueZlayBrpJfk/gKOA5SbNT2c+AXwOTJI0FXgcOAYiIFyRNInvwphI4OSKqnm0+iU9vVbo3bZAl52slzSPrcY5ObS2VdD7ZRDfAeRGxtK5gnTzNrDT53YZUr4iYQe1rL+1XS51xwLgaymcCO9VQvpKUfGvYNx4Yn2+8Tp5mVpLsPs/ye8LIydPMSqRCJoxaDSdPMyuZe55mZoVqoDHPlsbJ08xK4jFPM7MilWHudPI0s9K552lmVqjCnm1vNZw8zawkVet5lhsnTzMrUet7M2Y+nDzNrGRlmDudPM2sdO55mpkVSJ4wMjMrjnueZmZFKMPc6eRpZqVzz9PMrFBeGMTMrHDyfZ5mZsVp69l2M7PClWHH08nTzEqTvVa4/LKnk6eZlawMr9prT56S/gJEbfsj4geNEpGZtTjuea5tZpNFYWYtWhnmztqTZ0RMyP0uaYOIeL/xQzKzlkRA2zLMnm3qO0DSHpLmAC+m7wMkXdLokZlZy6DsPs98ttak3uQJ/AkYBiwBiIh/AXs3Ykxm1sJI+W2tSV6z7RHxRrX/aqxunHDMrKUR0Ka1ZcY85JM835D0RSAkdQB+QLqENzOD1terzEc+yfNE4M9AH+BNYApwcmMGZWYthxdDrkVELAaOaIJYzKyFKsfL9nxm2z8v6U5JiyQtlHSHpM83RXBm1jIoz63edqTxKc88n1N2rqQ3Jc1O24icfWdKmifpZUnDcsoHSXou7btQadJGUkdJE1P5E5K2zKkzRtIraRtTX6z5zLbfAEwCegObAX8HbsyjnpmViQa8VelqYHgN5RdExMC03ZPOuQMwGtgx1blEUtt0/KXA8UC/tFW1ORZYFhHbABcAv0ltdQfOAYYCQ4BzJHWrK9B8kqci4tqIqEzbddTx2KaZlZdstj2/rT4R8RCwNM9TjwRuiohVEfEaMA8YIqk30CUiHouIAK4BRuXUqXoA6GZgv9QrHQZMjYilEbEMmErNSXyNWpOnpO4pGz8g6QxJW0r6nKSfAHfn+ePMrLUr7Cb5npJm5mzH53mWUyQ9my7rq3qEfYA3co6pSGV90ufq5WvViYhKYDnQo462alXXhNHTZD3Mqv9enJCzL4Dz62rYzMpHAbPtiyNicIHNX0qWb6ryzh+AY6l5GDXqKKfIOjWq69n2reqqaGYGn162N5aIeHvNuaQrgLvS1wpg85xD+wJvpfK+NZTn1qmQ1A7oSjZMUAHsU63Og3XFlc+YJ5J2knSopKOrtnzqmVl5aMxn29MYZpUDgaqZ+MnA6DSDvhXZxNCTEbEAWCFp9zSeeTRwR06dqpn0g4HpaVx0CrC/pG5pWGD/VFareu/zlHQOWUbeAbgH+Dowg2wQ1swsr9uQ8mpHupEs3/SUVEE2A76PpIFkl9HzSUOIEfGCpEnAHKASODkiqh4dP4ls5r4TcG/aAK4ErpU0j6zHOTq1tVTS+cBT6bjzIqLOiat8njA6GBgAPBMRx0jaFPhbHvXMrAxIDXeTfER8p4biK+s4fhwwrobymcBONZSvBA6ppa3xwPh8Y80neX4YEZ9IqpTUBVgI+CZ5M1ujDB8wyit5zpS0MXAF2Qz8e8CTjRmUmbUsfra9BhHxvfTxMkn3kd18+mzjhmVmLYVQWT7bXtcL4Hata19EzGqckMysRWmFCx3no66e5x/q2BfAvg0cC7tsvwWPPHFRQzdrZkljXV23tlds5KOum+S/0pSBmFnLldcN461MXq/hMDOrjXDP08ysKO3KsOvp5GlmJcnejFl+Pc98VpKXpCMlnZ2+byFpSOOHZmYtRUOt59mS5NPZvgTYA6h6bGoFcHGjRWRmLY7f216zoRGxq6RnACJiWXoFsZmZ39teh4/Te0ECQNImwCeNGpWZtShtyy935pU8LwRuA3pJGke2ytJZjRqVmbUYkh/PrFFEXC/paWA/sh76qIh4sdEjM7MWowxzZ16LIW8BfADcmVsWEa83ZmBm1nK0tpn0fORz2X43n74g6TPAVsDLZO9KNrMy5wmjWkTEF3K/p9WWTqjlcDMrQ2WYOwt/wigiZknarTGCMbMWSNC2DLNnPmOeP8r52gbYFVjUaBGZWYvS2K8eXl/l0/PcKOdzJdkY6C2NE46ZtUROntWkm+M3jIgfN1E8ZtYClePCIHW9hqNdRFTW9ToOMzNftq/rSbLxzdmSJgN/B96v2hkRtzZybGbWErTCRT/ykc+YZ3dgCdk7i6ru9wzAydPMENCuDLuedSXPXmmm/Xk+TZpVolGjMrMWxT3PtbUFNmTtpFnFydPMEtGmxjTRutWVPBdExHlNFomZtUjZC+CaO4qmV1fyLMM/DjMrWCt8xUY+6kqe+zVZFGbWYgloW4bZs9bkGRFLmzIQM2u5ynFVpTJ827KZNbSGegGcpPGSFkp6Pqesu6Spkl5J/+yWs+9MSfMkvSxpWE75IEnPpX0XKj0CJamjpImp/AlJW+bUGZPO8YqkMfXF6uRpZiURWSLJZ8vD1cDwamVnANMioh8wLX1H0g7AaLK1hYcDl6RHygEuBY4H+qWtqs2xwLKI2Aa4APhNaqs7cA4wFBgCnJObpGvi5GlmpVH2bHs+W30i4iGg+pDhSGBC+jwBGJVTflNErIqI14B5wBBJvYEuEfFYRARwTbU6VW3dDOyXeqXDgKkRsTQilgFTWTeJr6Xg9TzNzKorYMSzp6SZOd8vj4jL66mzaUQsAIiIBZJ6pfI+wOM5x1Wkso/T5+rlVXXeSG1VSloO9Mgtr6FOjZw8zawkoqDFkBdHxOAGPHV11Z+GzC0vtk6NfNluZiVrqAmjWrydLsVJ/1yYyiuAzXOO6wu8lcr71lC+Vh1J7YCuZMMEtbVVKydPMytRfuOdJaz5ORmomv0eA9yRUz46zaBvRTYx9GS6xF8hafc0nnl0tTpVbR0MTE/jolOA/SV1SxNF+6eyWvmy3cxKUjXb3iBtSTcC+5CNjVaQzYD/GpgkaSzwOnAIQES8IGkSMIfsLRcnR8Tq1NRJZDP3nYB70wZwJXCtpHlkPc7Rqa2lks4HnkrHnVffve5OnmZWsoZaST4ivlPLrhqfeIyIccC4GspnAjvVUL6SlHxr2DceGJ9vrE6eZlay8nu+yMnTzEokv3rYzKw4fgGcmVkRyi91OnmaWQMow46nk6eZlSa7Van8sqeTp5mVzD1PM7OCqSwXQ3byNLOS+LLdzKwYpS360WI5eZpZyZw8zcyKoDK8bPeSdA3shOOOZYvNejFo4NprElxy0V/Yecf+7DpgR352xk+aKTrLtXr1anYfvAvfHnkAAEcefhhDBw1k6KCB9N9mS4YOGrjm2OeefZYvf2kPdh2wI4MHfoGVK1c2U9Trn6rFkPPZWhP3PBvYUWO+y4nfO4Xjjj16Tdk/H3yAu+68g6dmPUvHjh1ZuHBhHS1YU7nowj/Tf/vtWfHuuwBcd8PENft++uPT6Nq1KwCVlZUcO+ZIrrz6WnYeMIAlS5bQvn37Zol5fdXK8mJe3PNsYF/aa2+6d+++Vtnlf72U039yBh07dgSgV69eNVW1JlRRUcF9997NMccet86+iOCWmydx6GHZ6mj/mHo/O31hZ3YeMACAHj160LZt23XqlTPl+b/WxMmzCcybO5dHZjzMXl8cytf2/TIzn3qq/krWqH582qmM+9VvadNm3X8FHpnxMJv22pRt+vUD4JW5c5HEN0cMY4/dduUPv/9tU4e7XhPQRvltrUmjJc+aXl5fripXV7Js2TIeeuRxfvnr33Hk4YeSrfxvzeGeu++i1ya92HXQoBr3T7rpRg4Z/emavJWrK3n00Rlcdc31TPvnDCbffhsPTJ/WVOG2APn2O1tX9mzMnufV1PPe43LRp09fRh34bSSx25AhtGnThsWLFzd3WGXrsUcf4a67JtN/my05+ojRPPjAdI45+kggG9+84/ZbOfiQw9Yc36dPX/ba68v07NmTzp07M/zrI3jmmVnNFf76J8+Xv7W2cdFGS561vLy+LH3zW6N48IHpQHYJ+NFHH9GzZ89mjqp8nT/uV7w6v4KX583nmutvYp+v7MtV11wHwPRp/2Db/tvRt++nL1/82v7DeP65Z/nggw+orKzk4Yf+yfbb79Bc4a93PNveTCQdDxwPsPkWWzRzNKU7+sjv8PA/H2Tx4sVsvWVf/vfsXzDmmGM54bhjGTRwJzq078Dfxk8oy8VjW4K/T7xpzURRlW7duvGDU3/El/bYDUkMGz6Cr4/4RjNFuH4qx/83qzHH3iRtCdwVEeu8iKkmgwYNjkeemNlo8ZiVuz2HDubpp2c2aK7b/gu7xFW3P5DXsXts0+3piBjckOdvLs3e8zSzlq+1TQblw8nTzEpWjqNQjXmr0o3AY0B/SRXphfVm1gopz601abSeZx0vrzezVkT47ZlmZoVrhfdw5sPJ08xKVoa508nTzBpAGWZPJ08zK1Hre249H06eZlaSqlWVyo2XpDOz0jXQvUqS5kt6TtJsSTNTWXdJUyW9kv7ZLef4MyXNk/SypGE55YNSO/MkXah0O4CkjpImpvIn0lOQRXHyNLOSNfCSdF+JiIE5j3GeAUyLiH7AtPQdSTsAo4EdyVZwu0RS1SrVl5KtmdEvbVUrvI0FlkXENsAFwG+K/c1OnmZWskZekm4kMCF9ngCMyim/KSJWRcRrwDxgiKTeQJeIeCyyxTuuqVanqq2bgf1U5E2qTp5mVrIGfMIogPslPZ1WXAPYNCIWAKR/Vr3Hpg/wRk7dilTWJ32uXr5WnYioBJYDPfILbW2eMDKz0hT27GXPqrHM5PKIuDzn+54R8ZakXsBUSS/Vc+bqoo7yuuoUzMnTzEqSzbbnnT0X17UkXUS8lf65UNJtwBDgbUm9I2JBuiSvev1sBbB5TvW+wFupvG8N5bl1KiS1A7pS5KLtvmw3s5I1xGW7pA0kbVT1GdgfeB6YDIxJh40B7kifJwOj0wz6VmQTQ0+mS/sVknZP45lHV6tT1dbBwPQoclFj9zzNrHQNc5/npsBtaf6mHXBDRNwn6SlgUlqZ7XXgEICIeEHSJGAOUAmcHBGrU1snkb1HrRNwb9oArgSulTSPrMc5uthgnTzNrGQN8YRRRPwbGFBD+RJgv1rqjAPG1VA+E1jnDRYRsZKUfEvl5GlmJfOqSmZmRSjD3OnkaWal8WLIZmbF8GLIZmbFKcPc6eRpZg2gDLOnk6eZlciLIZuZFaxcF0N28jSz0jl5mpkVzpftZmZF8K1KZmZFKMPc6eRpZiXyTfJmZoXz45lmZkUqv9Tp5GlmDaAMO55OnmZWOt+qZGZWjPLLnU6eZla6MsydTp5mVhqpoFcPtxpOnmZWuvLLnU6eZla6MsydTp5mVroyvGp38jSzUnkxZDOzgmWPZzZ3FE3PydPMSubkaWZWBF+2m5kVykvSmZkVTvhWJTOz4pRh9nTyNLOS+fFMM7MilF/qdPI0s4ZQhtnTydPMSlaOtyopIpo7hjUkLQL+09xxNIKewOLmDsIK0lr/zj4XEZs0ZIOS7iP788rH4ogY3pDnby7rVfJsrSTNjIjBzR2H5c9/Z1afNs0dgJlZS+TkaWZWBCfPpnF5cwdgBfPfmdXJY55mZkVwz9PMrAhOnmZmRXDyNDMrgpNnI5HUX9IektpLatvc8Vh+/Hdl+fKEUSOQ9G3gl8CbaZsJXB0R7zZrYFYrSdtGxNz0uW1ErG7umGz95p5nA5PUHjgMGBsR+wF3AJsDP5HUpVmDsxpJOgCYLekGgIhY7R6o1cfJs3F0Afqlz7cBdwEdgMOlMlz4cD0maQPgFOBU4CNJ14ETqNXPybOBRcTHwB+Bb0vaKyI+AWYAs4EvNWdstq6IeB84FrgBOB34TG4Cbc7YbP3m5Nk4HgbuB46StHdErI6IG4DNgAHNG5pVFxFvRcR7EbEYOAHoVJVAJe0qabvmjdDWR17PsxFExEpJ1wMBnJn+5VsFbAosaNbgrE4RsUTSCcDvJL0EtAW+0sxh2XrIybORRMQySVcAc8h6MyuBIyPi7eaNzOoTEYslPQt8HfhaRFQ0d0y2/vGtSk0gTTxEGv+09ZykbsAk4LSIeLa547H1k5OnWQ0kfSYiVjZ3HLb+cvI0MyuCZ9vNzIrg5GlmVgQnTzOzIjh5mpkVwcmzBZG0WtJsSc9L+rukziW0dbWkg9Pnv0naoY5j95H0xSLOMV/SOu/zrq282jHvFXiucyWdXmiMZsVy8mxZPoyIgRGxE/ARcGLuzmIXsoiI4yJiTh2H7AMUnDzNWjMnz5brYWCb1Ct8IC2n9pyktpJ+J+kpSc+mRw1R5iJJcyTdDfSqakjSg5IGp8/DJc2S9C9J0yRtSZak/yf1eveStImkW9I5npK0Z6rbQ9L9kp6R9Feg3hWkJN0u6WlJL0g6vtq+P6RYpknaJJVtLem+VOdhP3duzcWPZ7ZAktqRPTp4XyoaAuwUEa+lBLQ8InaT1BF4RNL9wC5Af+ALZM/YzwHGV2t3E+AKYO/UVveIWCrpMuC9iPh9Ou4G4IKImCFpC2AKsD1wDjAjIs6T9A1grWRYi2PTOToBT0m6JSKWABsAsyLiNElnp7ZPIXsl8IkR8YqkocAlwL5F/DGalcTJs2XpJGl2+vwwcCXZ5fSTEfFaKt8f2LlqPBPoSra26N7AjWmZtbckTa+h/d2Bh6raioiltcTxVWCHnKVJu0jaKJ3j26nu3ZKW5fGbfiDpwPR58xTrEuATYGIqvw64VdKG6ff+PefcHfM4h1mDc/JsWT6MiIG5BSmJvJ9bBHw/IqZUO24E2SpPdVEex0A23LNHRHxYQyx5P7ImaR+yRLxHRHwg6UHgM7UcHum871T/MzBrDh7zbH2mACel14Egadu0WvpDwOg0JtqbmpdZewz4sqStUt3uqXwFsFHOcfeTXUKTjhuYPj4EHJHKvg50qyfWrsCylDi3I+v5VmkDVPWeDycbDngXeE3SIekckuT1Ua1ZOHm2Pn8jG8+cJel54K9kVxi3Aa8AzwGXAv+sXjEiFpGNU94q6V98etl8J3Bg1YQR8ANgcJqQmsOns/6/APaWNIts+OD1emK9D2iXln87H3g8Z9/7wI6SniYb0zwvlR8BjE3xvQCMzOPPxKzBeWEQM7MiuOdpZlYEJ08zsyI4eZqZFcHJ08ysCE6eZmZFcPI0MyuCk6eZWRH+P07d7DMhK/NGAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Confusion matrix on the whole dataset\n",
    "y_pred = model.predict(X)\n",
    "y_expected = pd.DataFrame(y)\n",
    "cnf_matrix = confusion_matrix(y_expected, y_pred.round())\n",
    "plot_confusion_matrix(cnf_matrix, classes=[0,1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = accuracy_score(y, y_pred.round())\n",
    "prec = precision_score(y, y_pred.round())\n",
    "rec = recall_score(y, y_pred.round())\n",
    "f1 = f1_score(y, y_pred.round())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999860</td>\n",
       "      <td>0.069106</td>\n",
       "      <td>0.930894</td>\n",
       "      <td>0.987069</td>\n",
       "      <td>0.958159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.999772</td>\n",
       "      <td>0.073171</td>\n",
       "      <td>0.926829</td>\n",
       "      <td>0.940206</td>\n",
       "      <td>0.933470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PlainNeuralNetwork</td>\n",
       "      <td>0.999403</td>\n",
       "      <td>0.211382</td>\n",
       "      <td>0.788618</td>\n",
       "      <td>0.854626</td>\n",
       "      <td>0.820296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>WeightedNeuralNetwork</td>\n",
       "      <td>0.976349</td>\n",
       "      <td>0.048780</td>\n",
       "      <td>0.951220</td>\n",
       "      <td>0.065181</td>\n",
       "      <td>0.122002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>UnderSampledNeuralNetwork</td>\n",
       "      <td>0.965801</td>\n",
       "      <td>0.032520</td>\n",
       "      <td>0.967480</td>\n",
       "      <td>0.046667</td>\n",
       "      <td>0.089039</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Model  Accuracy  FalseNegRate    Recall  Precision  \\\n",
       "0               RandomForest  0.999860      0.069106  0.930894   0.987069   \n",
       "1               DecisionTree  0.999772      0.073171  0.926829   0.940206   \n",
       "2         PlainNeuralNetwork  0.999403      0.211382  0.788618   0.854626   \n",
       "3      WeightedNeuralNetwork  0.976349      0.048780  0.951220   0.065181   \n",
       "4  UnderSampledNeuralNetwork  0.965801      0.032520  0.967480   0.046667   \n",
       "\n",
       "   F1 Score  \n",
       "0  0.958159  \n",
       "1  0.933470  \n",
       "2  0.820296  \n",
       "3  0.122002  \n",
       "4  0.089039  "
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_results = pd.DataFrame([['UnderSampledNeuralNetwork', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_fullset = results_fullset.append(model_results, ignore_index = True)\n",
    "results_fullset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Oversampling technique using SMOTE\n",
    "(Synthetic Minority Oversample TEchnique)\n",
    "- the alternative to down-sampling is the upsampling of the minority class to the same level as the dominant one.\n",
    "- This can be achieved with the SMOTE method where a new vector is generated between 2 existing datapoints. Applying this technique allows to massively increase the number of fraudulent transactions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "#%%bash\n",
    "#pip install -U imbalanced-learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "from imblearn.over_sampling import SMOTE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_resample, y_resample = SMOTE().fit_sample(X,y.values.ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 1, 1, 1], dtype=int64)"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_resample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of total transactions before SMOTE upsampling:  284807 ...after SMOTE upsampling:  568630\n",
      "Number of fraudulent transactions before SMOTE upsampling:  492 ...after SMOTE upsampling:  284315\n"
     ]
    }
   ],
   "source": [
    "print('Number of total transactions before SMOTE upsampling: ', len(y), '...after SMOTE upsampling: ', len(y_resample))\n",
    "print('Number of fraudulent transactions before SMOTE upsampling: ', len(y[y.Class==1]), \n",
    "      '...after SMOTE upsampling: ', np.sum(y_resample[y_resample==1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_resample = pd.DataFrame(y_resample)\n",
    "X_resample = pd.DataFrame(X_resample)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X_resample,y_resample,test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = np.array(X_train)\n",
    "X_test = np.array(X_test)\n",
    "y_train = np.array(y_train)\n",
    "y_test = np.array(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential([\n",
    "    Dense(units=16, input_dim = 29,activation='relu'),   # input of 29 columns as shown above\n",
    "    Dense(units=24,activation='relu'),\n",
    "    Dropout(0.5),\n",
    "    Dense(24,activation='relu'),\n",
    "    Dense(24,activation='relu'),\n",
    "    Dense(1,activation='sigmoid'),                        # binary classification fraudulent or not\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "398041/398041 [==============================] - 70s 177us/step - loss: 0.0442 - accuracy: 0.9841\n",
      "Epoch 2/5\n",
      "398041/398041 [==============================] - 68s 172us/step - loss: 0.0163 - accuracy: 0.9954\n",
      "Epoch 3/5\n",
      "398041/398041 [==============================] - 66s 167us/step - loss: 0.0126 - accuracy: 0.9966\n",
      "Epoch 4/5\n",
      "398041/398041 [==============================] - 69s 173us/step - loss: 0.0115 - accuracy: 0.9970\n",
      "Epoch 5/5\n",
      "398041/398041 [==============================] - 70s 175us/step - loss: 0.0107 - accuracy: 0.9974\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x19d43e022c8>"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.compile(optimizer='adam',loss='binary_crossentropy',metrics=['accuracy'])\n",
    "model.fit(X_train,y_train,batch_size=15,epochs=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[85150   232]\n",
      " [    0 85207]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = model.predict(X_test)\n",
    "y_expected = pd.DataFrame(y_test)\n",
    "cnf_matrix = confusion_matrix(y_expected, y_pred.round())\n",
    "plot_confusion_matrix(cnf_matrix, classes=[0,1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Notice the absence of False Negatives. The model is able to detect all fraudulent transactions on the unseen test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = accuracy_score(y_test, y_pred.round())\n",
    "prec = precision_score(y_test, y_pred.round())\n",
    "rec = recall_score(y_test, y_pred.round())\n",
    "f1 = f1_score(y_test, y_pred.round())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999544</td>\n",
       "      <td>0.224490</td>\n",
       "      <td>0.775510</td>\n",
       "      <td>0.950000</td>\n",
       "      <td>0.853933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.999239</td>\n",
       "      <td>0.244898</td>\n",
       "      <td>0.755102</td>\n",
       "      <td>0.792857</td>\n",
       "      <td>0.773519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PlainNeuralNetwork</td>\n",
       "      <td>0.999380</td>\n",
       "      <td>0.217687</td>\n",
       "      <td>0.782313</td>\n",
       "      <td>0.845588</td>\n",
       "      <td>0.812721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>WeightedNeuralNetwork</td>\n",
       "      <td>0.975937</td>\n",
       "      <td>0.102041</td>\n",
       "      <td>0.897959</td>\n",
       "      <td>0.060746</td>\n",
       "      <td>0.113793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>UnderSampledNeuralNetwork</td>\n",
       "      <td>0.959459</td>\n",
       "      <td>0.040268</td>\n",
       "      <td>0.959732</td>\n",
       "      <td>0.959732</td>\n",
       "      <td>0.959732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>OverSampledNeuralNetwork</td>\n",
       "      <td>0.998640</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.997285</td>\n",
       "      <td>0.998640</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Model  Accuracy  FalseNegRate    Recall  Precision  \\\n",
       "0               RandomForest  0.999544      0.224490  0.775510   0.950000   \n",
       "1               DecisionTree  0.999239      0.244898  0.755102   0.792857   \n",
       "2         PlainNeuralNetwork  0.999380      0.217687  0.782313   0.845588   \n",
       "3      WeightedNeuralNetwork  0.975937      0.102041  0.897959   0.060746   \n",
       "4  UnderSampledNeuralNetwork  0.959459      0.040268  0.959732   0.959732   \n",
       "5   OverSampledNeuralNetwork  0.998640      0.000000  1.000000   0.997285   \n",
       "\n",
       "   F1 Score  \n",
       "0  0.853933  \n",
       "1  0.773519  \n",
       "2  0.812721  \n",
       "3  0.113793  \n",
       "4  0.959732  \n",
       "5  0.998640  "
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Store results in dataframe for comparing various Models\n",
    "model_results = pd.DataFrame([['OverSampledNeuralNetwork', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_testset = results_testset.append(model_results, ignore_index = True)\n",
    "results_testset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All metrics are excellent for this last model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[283561    754]\n",
      " [     0    492]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Confusion matrix on the whole dataset\n",
    "y_pred = model.predict(X)\n",
    "y_expected = pd.DataFrame(y)\n",
    "cnf_matrix = confusion_matrix(y_expected, y_pred.round())\n",
    "plot_confusion_matrix(cnf_matrix, classes=[0,1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Notice the absence of False Negatives. The model is able to detect all fraudulent transactions on the full dataset.\n",
    "- Notice the limited number of False Positives which means a lot less verification work (on legitimate transactions) for the fraud departement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = accuracy_score(y, y_pred.round())\n",
    "prec = precision_score(y, y_pred.round())\n",
    "rec = recall_score(y, y_pred.round())\n",
    "f1 = f1_score(y, y_pred.round())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "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>Model</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>FalseNegRate</th>\n",
       "      <th>Recall</th>\n",
       "      <th>Precision</th>\n",
       "      <th>F1 Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.999860</td>\n",
       "      <td>0.069106</td>\n",
       "      <td>0.930894</td>\n",
       "      <td>0.987069</td>\n",
       "      <td>0.958159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.999772</td>\n",
       "      <td>0.073171</td>\n",
       "      <td>0.926829</td>\n",
       "      <td>0.940206</td>\n",
       "      <td>0.933470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PlainNeuralNetwork</td>\n",
       "      <td>0.999403</td>\n",
       "      <td>0.211382</td>\n",
       "      <td>0.788618</td>\n",
       "      <td>0.854626</td>\n",
       "      <td>0.820296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>WeightedNeuralNetwork</td>\n",
       "      <td>0.976349</td>\n",
       "      <td>0.048780</td>\n",
       "      <td>0.951220</td>\n",
       "      <td>0.065181</td>\n",
       "      <td>0.122002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>UnderSampledNeuralNetwork</td>\n",
       "      <td>0.965801</td>\n",
       "      <td>0.032520</td>\n",
       "      <td>0.967480</td>\n",
       "      <td>0.046667</td>\n",
       "      <td>0.089039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>OverSampledNeuralNetwork</td>\n",
       "      <td>0.997353</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.394864</td>\n",
       "      <td>0.566168</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Model  Accuracy  FalseNegRate    Recall  Precision  \\\n",
       "0               RandomForest  0.999860      0.069106  0.930894   0.987069   \n",
       "1               DecisionTree  0.999772      0.073171  0.926829   0.940206   \n",
       "2         PlainNeuralNetwork  0.999403      0.211382  0.788618   0.854626   \n",
       "3      WeightedNeuralNetwork  0.976349      0.048780  0.951220   0.065181   \n",
       "4  UnderSampledNeuralNetwork  0.965801      0.032520  0.967480   0.046667   \n",
       "5   OverSampledNeuralNetwork  0.997353      0.000000  1.000000   0.394864   \n",
       "\n",
       "   F1 Score  \n",
       "0  0.958159  \n",
       "1  0.933470  \n",
       "2  0.820296  \n",
       "3  0.122002  \n",
       "4  0.089039  \n",
       "5  0.566168  "
      ]
     },
     "execution_count": 133,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_results = pd.DataFrame([['OverSampledNeuralNetwork', acc, 1-rec, rec, prec, f1]],\n",
    "               columns = ['Model', 'Accuracy', 'FalseNegRate', 'Recall', 'Precision', 'F1 Score'])\n",
    "results_fullset = results_fullset.append(model_results, ignore_index = True)\n",
    "results_fullset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "tensorflow_py37",
   "language": "python",
   "name": "tensorflow_py37"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}