Deep Learning/Self Organizing Maps/Credit Fraud Detection with SOM.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "10ba0f18",
   "metadata": {},
   "source": [
    "# Preparing the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6ec5cc46",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>A1</th>\n",
       "      <th>A2</th>\n",
       "      <th>A3</th>\n",
       "      <th>A4</th>\n",
       "      <th>A5</th>\n",
       "      <th>A6</th>\n",
       "      <th>A7</th>\n",
       "      <th>A8</th>\n",
       "      <th>A9</th>\n",
       "      <th>A10</th>\n",
       "      <th>A11</th>\n",
       "      <th>A12</th>\n",
       "      <th>A13</th>\n",
       "      <th>A14</th>\n",
       "      <th>Class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>15776156</td>\n",
       "      <td>1</td>\n",
       "      <td>22.08</td>\n",
       "      <td>11.46</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1.585</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>100</td>\n",
       "      <td>1213</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>15739548</td>\n",
       "      <td>0</td>\n",
       "      <td>22.67</td>\n",
       "      <td>7.00</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "      <td>0.165</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>160</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>15662854</td>\n",
       "      <td>0</td>\n",
       "      <td>29.58</td>\n",
       "      <td>1.75</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1.250</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>280</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>15687688</td>\n",
       "      <td>0</td>\n",
       "      <td>21.67</td>\n",
       "      <td>11.50</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>15715750</td>\n",
       "      <td>1</td>\n",
       "      <td>20.17</td>\n",
       "      <td>8.17</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>1.960</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>60</td>\n",
       "      <td>159</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   CustomerID  A1     A2     A3  A4  A5  A6     A7  A8  A9  A10  A11  A12  \\\n",
       "0    15776156   1  22.08  11.46   2   4   4  1.585   0   0    0    1    2   \n",
       "1    15739548   0  22.67   7.00   2   8   4  0.165   0   0    0    0    2   \n",
       "2    15662854   0  29.58   1.75   1   4   4  1.250   0   0    0    1    2   \n",
       "3    15687688   0  21.67  11.50   1   5   3  0.000   1   1   11    1    2   \n",
       "4    15715750   1  20.17   8.17   2   6   4  1.960   1   1   14    0    2   \n",
       "\n",
       "   A13   A14  Class  \n",
       "0  100  1213      0  \n",
       "1  160     1      0  \n",
       "2  280     1      0  \n",
       "3    0     1      1  \n",
       "4   60   159      1  "
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Importing the libraries\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Loading the data\n",
    "# https://archive.ics.uci.edu/ml/datasets/statlog+(australian+credit+approval)\n",
    "df = pd.read_csv(\"Credit_Card_Applications.csv\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4ce54a76",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split the data\n",
    "X = df.iloc[:, :-1].values\n",
    "y = df.iloc[:, -1].values # Class if credit application is approved or not\n",
    "\n",
    "# Feature Scaling\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "mms = MinMaxScaler(feature_range = (0, 1))\n",
    "X = mms.fit_transform(X)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8db96dcd",
   "metadata": {},
   "source": [
    "# Self Organizing Map model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3b43a46f",
   "metadata": {},
   "source": [
    "## Training the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b1d96884",
   "metadata": {},
   "outputs": [],
   "source": [
    "from minisom import MiniSom\n",
    "# Initiliaze the SOM model\n",
    "som = MiniSom(x = 10, y = 10, input_len = X.shape[1],\n",
    "             sigma = 1.0, learning_rate = 0.5,\n",
    "             random_seed = 123) # (x, y) = SOM output grid dimension\n",
    "# Initiliaze random weights for the model\n",
    "som.random_weights_init(data = X)\n",
    "# Train the SOM model\n",
    "som.train_random(data = X, num_iteration = 500)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ac4fe70",
   "metadata": {},
   "source": [
    "## Visualising the results\n",
    "Depending on the winning node with a high Mean Interneuron Distance (MID), we will identify potential fraud cases. After visualising the SOM below, the outliers with white color and high MID and also got the approval of the bank represent the potential fraud cases."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "671dc5ca",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (12, 10))\n",
    "plt.bone() # Initiate the frame\n",
    "plt.pcolor(som.distance_map().T) # Color the distance matrix\n",
    "plt.colorbar() # Legend for colors\n",
    "markers = ['o', 's'] # o : circle, s: square\n",
    "colors = ['r', 'g'] # Red = rejected, Green = approved\n",
    "for i, x in enumerate(X):\n",
    "    w = som.winner(x) # Take the customer's winning node coordinates\n",
    "    plt.plot(w[0] + 0.5, # + 0.5 for centrering the marker\n",
    "             w[1] + 0.5,\n",
    "             markers[y[i]],\n",
    "             markeredgecolor = colors[y[i]],\n",
    "             markerfacecolor = 'None',\n",
    "             markeredgewidth = 1.5,\n",
    "             markersize = 20)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33589b72",
   "metadata": {},
   "source": [
    "## Finding the potential frauds\n",
    "According the mapping above, the outliers with high MIDs at (5, 6), (4, 7) and (5, 8) coordinates include potential fraud cases."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6a23d277",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "</style><table id=\"T_dfe08_\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >Fraud Customer ID</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_dfe08_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "                        <td id=\"T_dfe08_row0_col0\" class=\"data row0 col0\" >15597536</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "                        <td id=\"T_dfe08_row1_col0\" class=\"data row1 col0\" >15571284</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "                        <td id=\"T_dfe08_row2_col0\" class=\"data row2 col0\" >15570947</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "                        <td id=\"T_dfe08_row3_col0\" class=\"data row3 col0\" >15641715</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "                        <td id=\"T_dfe08_row4_col0\" class=\"data row4 col0\" >15589881</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "                        <td id=\"T_dfe08_row5_col0\" class=\"data row5 col0\" >15592326</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "                        <td id=\"T_dfe08_row6_col0\" class=\"data row6 col0\" >15580579</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "                        <td id=\"T_dfe08_row7_col0\" class=\"data row7 col0\" >15665374</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "                        <td id=\"T_dfe08_row8_col0\" class=\"data row8 col0\" >15731586</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "                        <td id=\"T_dfe08_row9_col0\" class=\"data row9 col0\" >15575605</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row10\" class=\"row_heading level0 row10\" >10</th>\n",
       "                        <td id=\"T_dfe08_row10_col0\" class=\"data row10 col0\" >15644453</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row11\" class=\"row_heading level0 row11\" >11</th>\n",
       "                        <td id=\"T_dfe08_row11_col0\" class=\"data row11 col0\" >15646594</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row12\" class=\"row_heading level0 row12\" >12</th>\n",
       "                        <td id=\"T_dfe08_row12_col0\" class=\"data row12 col0\" >15699238</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row13\" class=\"row_heading level0 row13\" >13</th>\n",
       "                        <td id=\"T_dfe08_row13_col0\" class=\"data row13 col0\" >15577771</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row14\" class=\"row_heading level0 row14\" >14</th>\n",
       "                        <td id=\"T_dfe08_row14_col0\" class=\"data row14 col0\" >15599272</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row15\" class=\"row_heading level0 row15\" >15</th>\n",
       "                        <td id=\"T_dfe08_row15_col0\" class=\"data row15 col0\" >15711299</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row16\" class=\"row_heading level0 row16\" >16</th>\n",
       "                        <td id=\"T_dfe08_row16_col0\" class=\"data row16 col0\" >15666096</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row17\" class=\"row_heading level0 row17\" >17</th>\n",
       "                        <td id=\"T_dfe08_row17_col0\" class=\"data row17 col0\" >15605791</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row18\" class=\"row_heading level0 row18\" >18</th>\n",
       "                        <td id=\"T_dfe08_row18_col0\" class=\"data row18 col0\" >15571415</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row19\" class=\"row_heading level0 row19\" >19</th>\n",
       "                        <td id=\"T_dfe08_row19_col0\" class=\"data row19 col0\" >15565714</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row20\" class=\"row_heading level0 row20\" >20</th>\n",
       "                        <td id=\"T_dfe08_row20_col0\" class=\"data row20 col0\" >15655464</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row21\" class=\"row_heading level0 row21\" >21</th>\n",
       "                        <td id=\"T_dfe08_row21_col0\" class=\"data row21 col0\" >15608688</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row22\" class=\"row_heading level0 row22\" >22</th>\n",
       "                        <td id=\"T_dfe08_row22_col0\" class=\"data row22 col0\" >15650313</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row23\" class=\"row_heading level0 row23\" >23</th>\n",
       "                        <td id=\"T_dfe08_row23_col0\" class=\"data row23 col0\" >15575438</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row24\" class=\"row_heading level0 row24\" >24</th>\n",
       "                        <td id=\"T_dfe08_row24_col0\" class=\"data row24 col0\" >15565996</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row25\" class=\"row_heading level0 row25\" >25</th>\n",
       "                        <td id=\"T_dfe08_row25_col0\" class=\"data row25 col0\" >15644400</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row26\" class=\"row_heading level0 row26\" >26</th>\n",
       "                        <td id=\"T_dfe08_row26_col0\" class=\"data row26 col0\" >15592914</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row27\" class=\"row_heading level0 row27\" >27</th>\n",
       "                        <td id=\"T_dfe08_row27_col0\" class=\"data row27 col0\" >15581871</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row28\" class=\"row_heading level0 row28\" >28</th>\n",
       "                        <td id=\"T_dfe08_row28_col0\" class=\"data row28 col0\" >15611973</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row29\" class=\"row_heading level0 row29\" >29</th>\n",
       "                        <td id=\"T_dfe08_row29_col0\" class=\"data row29 col0\" >15593178</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row30\" class=\"row_heading level0 row30\" >30</th>\n",
       "                        <td id=\"T_dfe08_row30_col0\" class=\"data row30 col0\" >15573077</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row31\" class=\"row_heading level0 row31\" >31</th>\n",
       "                        <td id=\"T_dfe08_row31_col0\" class=\"data row31 col0\" >15645571</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row32\" class=\"row_heading level0 row32\" >32</th>\n",
       "                        <td id=\"T_dfe08_row32_col0\" class=\"data row32 col0\" >15638272</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row33\" class=\"row_heading level0 row33\" >33</th>\n",
       "                        <td id=\"T_dfe08_row33_col0\" class=\"data row33 col0\" >15652289</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row34\" class=\"row_heading level0 row34\" >34</th>\n",
       "                        <td id=\"T_dfe08_row34_col0\" class=\"data row34 col0\" >15642001</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row35\" class=\"row_heading level0 row35\" >35</th>\n",
       "                        <td id=\"T_dfe08_row35_col0\" class=\"data row35 col0\" >15620103</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row36\" class=\"row_heading level0 row36\" >36</th>\n",
       "                        <td id=\"T_dfe08_row36_col0\" class=\"data row36 col0\" >15642063</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row37\" class=\"row_heading level0 row37\" >37</th>\n",
       "                        <td id=\"T_dfe08_row37_col0\" class=\"data row37 col0\" >15588019</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row38\" class=\"row_heading level0 row38\" >38</th>\n",
       "                        <td id=\"T_dfe08_row38_col0\" class=\"data row38 col0\" >15599535</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row39\" class=\"row_heading level0 row39\" >39</th>\n",
       "                        <td id=\"T_dfe08_row39_col0\" class=\"data row39 col0\" >15678779</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row40\" class=\"row_heading level0 row40\" >40</th>\n",
       "                        <td id=\"T_dfe08_row40_col0\" class=\"data row40 col0\" >15654760</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row41\" class=\"row_heading level0 row41\" >41</th>\n",
       "                        <td id=\"T_dfe08_row41_col0\" class=\"data row41 col0\" >15575146</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_dfe08_level0_row42\" class=\"row_heading level0 row42\" >42</th>\n",
       "                        <td id=\"T_dfe08_row42_col0\" class=\"data row42 col0\" >15611794</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x271811f0280>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mappings = som.win_map(X)\n",
    "frauds = np.concatenate((mappings[(5, 6)], mappings[(4, 7)], mappings[(5, 8)]))\n",
    "frauds = mms.inverse_transform(frauds)\n",
    "results = pd.DataFrame([int(i) for i in frauds[:, 0]], \n",
    "                      columns = ['Fraud Customer ID'])\n",
    "results.style"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "888bcd11",
   "metadata": {},
   "source": [
    "# Building Artificial Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "66332162",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating the matrix of features\n",
    "customers = df.iloc[:, 1:].values\n",
    "\n",
    "# Creating the dependent variable\n",
    "is_fraud = np.zeros(df.shape[0], dtype = np.int64)\n",
    "for i in range(df.shape[0]):\n",
    "    if df.iloc[i, 0] in frauds:\n",
    "        is_fraud[i] = 1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "5a4b2c55",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Feature Scaling\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "sc = StandardScaler()\n",
    "customers = sc.fit_transform(customers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8f92fd97",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import tensorflow.keras.layers as F\n",
    "\n",
    "def build_ann():\n",
    "    model = tf.keras.models.Sequential()\n",
    "    model.add(F.Dense(units = 2, kernel_initializer = 'uniform', activation = 'relu', input_dim = customers.shape[1]))\n",
    "    model.add(F.Dense(units = 1, kernel_initializer = 'uniform', activation = 'sigmoid'))\n",
    "    model.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = ['accuracy'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7e25bebc",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/3\n",
      "690/690 [==============================] - 1s 715us/step - loss: 0.5062 - accuracy: 0.9304\n",
      "Epoch 2/3\n",
      "690/690 [==============================] - 0s 714us/step - loss: 0.2475 - accuracy: 0.9377\n",
      "Epoch 3/3\n",
      "690/690 [==============================] - 0s 718us/step - loss: 0.2064 - accuracy: 0.9377\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x27191451820>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Training the ANN model \n",
    "ann = build_ann()\n",
    "ann.fit(customers, is_fraud, batch_size = 1, epochs = 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "4d4e368a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get the predictions and rank their probabilities\n",
    "y_pred = ann.predict(customers)\n",
    "y_pred = np.concatenate((df.iloc[:, 0:1].values, y_pred), axis = 1)\n",
    "y_pred = y_pred[y_pred[:, 1].argsort()[::-1]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "64af59ba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "</style><table id=\"T_43ce6_\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >CustomerID</th>        <th class=\"col_heading level0 col1\" >Fraud Probability</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_43ce6_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "                        <td id=\"T_43ce6_row0_col0\" class=\"data row0 col0\" >15699238</td>\n",
       "                        <td id=\"T_43ce6_row0_col1\" class=\"data row0 col1\" >0.321697</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_43ce6_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "                        <td id=\"T_43ce6_row1_col0\" class=\"data row1 col0\" >15636626</td>\n",
       "                        <td id=\"T_43ce6_row1_col1\" class=\"data row1 col1\" >0.309194</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_43ce6_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "                        <td id=\"T_43ce6_row2_col0\" class=\"data row2 col0\" >15600027</td>\n",
       "                        <td id=\"T_43ce6_row2_col1\" class=\"data row2 col1\" >0.301901</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_43ce6_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "                        <td id=\"T_43ce6_row3_col0\" class=\"data row3 col0\" >15609070</td>\n",
       "                        <td id=\"T_43ce6_row3_col1\" class=\"data row3 col1\" >0.282320</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_43ce6_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "                        <td id=\"T_43ce6_row4_col0\" class=\"data row4 col0\" >15744044</td>\n",
       "                        <td id=\"T_43ce6_row4_col1\" class=\"data row4 col1\" >0.279483</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_43ce6_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "                        <td id=\"T_43ce6_row5_col0\" class=\"data row5 col0\" >15575605</td>\n",
       "                        <td id=\"T_43ce6_row5_col1\" class=\"data row5 col1\" >0.269781</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_43ce6_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "                        <td id=\"T_43ce6_row6_col0\" class=\"data row6 col0\" >15700549</td>\n",
       "                        <td id=\"T_43ce6_row6_col1\" class=\"data row6 col1\" >0.268856</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_43ce6_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "                        <td id=\"T_43ce6_row7_col0\" class=\"data row7 col0\" >15599272</td>\n",
       "                        <td id=\"T_43ce6_row7_col1\" class=\"data row7 col1\" >0.259603</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_43ce6_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "                        <td id=\"T_43ce6_row8_col0\" class=\"data row8 col0\" >15726167</td>\n",
       "                        <td id=\"T_43ce6_row8_col1\" class=\"data row8 col1\" >0.256724</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_43ce6_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "                        <td id=\"T_43ce6_row9_col0\" class=\"data row9 col0\" >15649101</td>\n",
       "                        <td id=\"T_43ce6_row9_col1\" class=\"data row9 col1\" >0.251475</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27193ceaac0>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "probs_per_customer = pd.DataFrame(y_pred, columns = ['CustomerID', 'Fraud Probability'])\n",
    "probs_per_customer['CustomerID'] = probs_per_customer['CustomerID'].astype(np.int64)\n",
    "probs_per_customer.head(10).style"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}