fraud-prediction.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "154bf8d0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:23.006353Z",
     "iopub.status.busy": "2022-10-07T11:58:23.004087Z",
     "iopub.status.idle": "2022-10-07T11:58:28.052857Z",
     "shell.execute_reply": "2022-10-07T11:58:28.051821Z"
    },
    "papermill": {
     "duration": 5.070185,
     "end_time": "2022-10-07T11:58:28.055849",
     "exception": false,
     "start_time": "2022-10-07T11:58:22.985664",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "50134ce8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:28.071957Z",
     "iopub.status.busy": "2022-10-07T11:58:28.071465Z",
     "iopub.status.idle": "2022-10-07T11:58:28.457129Z",
     "shell.execute_reply": "2022-10-07T11:58:28.456241Z"
    },
    "papermill": {
     "duration": 0.395694,
     "end_time": "2022-10-07T11:58:28.459314",
     "exception": false,
     "start_time": "2022-10-07T11:58:28.063620",
     "status": "completed"
    },
    "tags": []
   },
   "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>teacher_id</th>\n",
       "      <th>total_call_before_pt</th>\n",
       "      <th>total_call_after_pt</th>\n",
       "      <th>call_before_pt_yes</th>\n",
       "      <th>call_after_pt_yes</th>\n",
       "      <th>not_held</th>\n",
       "      <th>held_at_1</th>\n",
       "      <th>held_at_2</th>\n",
       "      <th>held_at_3</th>\n",
       "      <th>held_at_4</th>\n",
       "      <th>total_held</th>\n",
       "      <th>buy_premium_y</th>\n",
       "      <th>buy_premium_n</th>\n",
       "      <th>total_student</th>\n",
       "      <th>rate_not_held</th>\n",
       "      <th>rate_bp_y</th>\n",
       "      <th>rate_bp_n</th>\n",
       "      <th>fraud_possibility</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9</td>\n",
       "      <td>128</td>\n",
       "      <td>152</td>\n",
       "      <td>55</td>\n",
       "      <td>80</td>\n",
       "      <td>18</td>\n",
       "      <td>61</td>\n",
       "      <td>30</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>100</td>\n",
       "      <td>12</td>\n",
       "      <td>78</td>\n",
       "      <td>90</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.133333</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>Medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>16</td>\n",
       "      <td>9</td>\n",
       "      <td>19</td>\n",
       "      <td>5</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "      <td>0.181818</td>\n",
       "      <td>0.090909</td>\n",
       "      <td>0.909091</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>36</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>37</td>\n",
       "      <td>41</td>\n",
       "      <td>29</td>\n",
       "      <td>24</td>\n",
       "      <td>13</td>\n",
       "      <td>15</td>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>2</td>\n",
       "      <td>24</td>\n",
       "      <td>26</td>\n",
       "      <td>0.576923</td>\n",
       "      <td>0.076923</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>39</td>\n",
       "      <td>134</td>\n",
       "      <td>162</td>\n",
       "      <td>82</td>\n",
       "      <td>65</td>\n",
       "      <td>44</td>\n",
       "      <td>54</td>\n",
       "      <td>24</td>\n",
       "      <td>13</td>\n",
       "      <td>5</td>\n",
       "      <td>96</td>\n",
       "      <td>13</td>\n",
       "      <td>89</td>\n",
       "      <td>102</td>\n",
       "      <td>0.431373</td>\n",
       "      <td>0.127451</td>\n",
       "      <td>0.872549</td>\n",
       "      <td>Medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>325</th>\n",
       "      <td>1664</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>326</th>\n",
       "      <td>1670</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>327</th>\n",
       "      <td>1682</td>\n",
       "      <td>15</td>\n",
       "      <td>19</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>1691</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>329</th>\n",
       "      <td>1697</td>\n",
       "      <td>42</td>\n",
       "      <td>16</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>330 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     teacher_id  total_call_before_pt  total_call_after_pt  \\\n",
       "0             9                   128                  152   \n",
       "1            16                     9                   19   \n",
       "2            36                     3                    2   \n",
       "3            37                    41                   29   \n",
       "4            39                   134                  162   \n",
       "..          ...                   ...                  ...   \n",
       "325        1664                     6                    5   \n",
       "326        1670                     6                    4   \n",
       "327        1682                    15                   19   \n",
       "328        1691                     5                    7   \n",
       "329        1697                    42                   16   \n",
       "\n",
       "     call_before_pt_yes  call_after_pt_yes  not_held  held_at_1  held_at_2  \\\n",
       "0                    55                 80        18         61         30   \n",
       "1                     5                  9         2          8          2   \n",
       "2                     2                  2         0          2          0   \n",
       "3                    24                 13        15         10          3   \n",
       "4                    82                 65        44         54         24   \n",
       "..                  ...                ...       ...        ...        ...   \n",
       "325                   5                  5         0          5          0   \n",
       "326                   4                  4         2          2          0   \n",
       "327                  10                  9         2          8          2   \n",
       "328                   4                  5         1          3          1   \n",
       "329                   9                 10         0         10          1   \n",
       "\n",
       "     held_at_3  held_at_4  total_held  buy_premium_y  buy_premium_n  \\\n",
       "0            7          2         100             12             78   \n",
       "1            0          0          10              1             10   \n",
       "2            0          0           2              0              2   \n",
       "3            0          2          15              2             24   \n",
       "4           13          5          96             13             89   \n",
       "..         ...        ...         ...            ...            ...   \n",
       "325          0          0           5              0              5   \n",
       "326          0          0           2              0              4   \n",
       "327          0          0          10              2              8   \n",
       "328          0          0           4              0              5   \n",
       "329          1          0          12              4              6   \n",
       "\n",
       "     total_student  rate_not_held  rate_bp_y  rate_bp_n fraud_possibility  \n",
       "0               90       0.200000   0.133333   0.866667            Medium  \n",
       "1               11       0.181818   0.090909   0.909091               Low  \n",
       "2                2       0.000000   0.000000   1.000000               Low  \n",
       "3               26       0.576923   0.076923   0.923077               Low  \n",
       "4              102       0.431373   0.127451   0.872549            Medium  \n",
       "..             ...            ...        ...        ...               ...  \n",
       "325              5       0.000000   0.000000   1.000000               Low  \n",
       "326              4       0.500000   0.000000   1.000000               Low  \n",
       "327             10       0.200000   0.200000   0.800000               Low  \n",
       "328              5       0.200000   0.000000   1.000000               Low  \n",
       "329             10       0.000000   0.400000   0.600000               Low  \n",
       "\n",
       "[330 rows x 18 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_excel('../input/teacher-placement/placement_teacher.xlsx')\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "eb7052c3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:28.474884Z",
     "iopub.status.busy": "2022-10-07T11:58:28.473270Z",
     "iopub.status.idle": "2022-10-07T11:58:28.496525Z",
     "shell.execute_reply": "2022-10-07T11:58:28.495047Z"
    },
    "papermill": {
     "duration": 0.032954,
     "end_time": "2022-10-07T11:58:28.498826",
     "exception": false,
     "start_time": "2022-10-07T11:58:28.465872",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 330 entries, 0 to 329\n",
      "Data columns (total 18 columns):\n",
      " #   Column                Non-Null Count  Dtype  \n",
      "---  ------                --------------  -----  \n",
      " 0   teacher_id            330 non-null    int64  \n",
      " 1   total_call_before_pt  330 non-null    int64  \n",
      " 2   total_call_after_pt   330 non-null    int64  \n",
      " 3   call_before_pt_yes    330 non-null    int64  \n",
      " 4   call_after_pt_yes     330 non-null    int64  \n",
      " 5   not_held              330 non-null    int64  \n",
      " 6   held_at_1             330 non-null    int64  \n",
      " 7   held_at_2             330 non-null    int64  \n",
      " 8   held_at_3             330 non-null    int64  \n",
      " 9   held_at_4             330 non-null    int64  \n",
      " 10  total_held            330 non-null    int64  \n",
      " 11  buy_premium_y         330 non-null    int64  \n",
      " 12  buy_premium_n         330 non-null    int64  \n",
      " 13  total_student         330 non-null    int64  \n",
      " 14  rate_not_held         330 non-null    float64\n",
      " 15  rate_bp_y             330 non-null    float64\n",
      " 16  rate_bp_n             330 non-null    float64\n",
      " 17  fraud_possibility     330 non-null    object \n",
      "dtypes: float64(3), int64(14), object(1)\n",
      "memory usage: 46.5+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65cb1c11",
   "metadata": {
    "papermill": {
     "duration": 0.006055,
     "end_time": "2022-10-07T11:58:28.511224",
     "exception": false,
     "start_time": "2022-10-07T11:58:28.505169",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Correlation between columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "b2ed2459",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:28.525359Z",
     "iopub.status.busy": "2022-10-07T11:58:28.524535Z",
     "iopub.status.idle": "2022-10-07T11:58:28.544595Z",
     "shell.execute_reply": "2022-10-07T11:58:28.543630Z"
    },
    "papermill": {
     "duration": 0.029246,
     "end_time": "2022-10-07T11:58:28.546650",
     "exception": false,
     "start_time": "2022-10-07T11:58:28.517404",
     "status": "completed"
    },
    "tags": []
   },
   "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>teacher_id</th>\n",
       "      <th>total_call_before_pt</th>\n",
       "      <th>total_call_after_pt</th>\n",
       "      <th>call_before_pt_yes</th>\n",
       "      <th>call_after_pt_yes</th>\n",
       "      <th>not_held</th>\n",
       "      <th>held_at_1</th>\n",
       "      <th>held_at_2</th>\n",
       "      <th>held_at_3</th>\n",
       "      <th>held_at_4</th>\n",
       "      <th>total_held</th>\n",
       "      <th>buy_premium_y</th>\n",
       "      <th>buy_premium_n</th>\n",
       "      <th>total_student</th>\n",
       "      <th>rate_not_held</th>\n",
       "      <th>rate_bp_y</th>\n",
       "      <th>rate_bp_n</th>\n",
       "      <th>fraud_possibility</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9</td>\n",
       "      <td>128</td>\n",
       "      <td>152</td>\n",
       "      <td>55</td>\n",
       "      <td>80</td>\n",
       "      <td>18</td>\n",
       "      <td>61</td>\n",
       "      <td>30</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>100</td>\n",
       "      <td>12</td>\n",
       "      <td>78</td>\n",
       "      <td>90</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.133333</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>16</td>\n",
       "      <td>9</td>\n",
       "      <td>19</td>\n",
       "      <td>5</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "      <td>0.181818</td>\n",
       "      <td>0.090909</td>\n",
       "      <td>0.909091</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>36</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>37</td>\n",
       "      <td>41</td>\n",
       "      <td>29</td>\n",
       "      <td>24</td>\n",
       "      <td>13</td>\n",
       "      <td>15</td>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>2</td>\n",
       "      <td>24</td>\n",
       "      <td>26</td>\n",
       "      <td>0.576923</td>\n",
       "      <td>0.076923</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>39</td>\n",
       "      <td>134</td>\n",
       "      <td>162</td>\n",
       "      <td>82</td>\n",
       "      <td>65</td>\n",
       "      <td>44</td>\n",
       "      <td>54</td>\n",
       "      <td>24</td>\n",
       "      <td>13</td>\n",
       "      <td>5</td>\n",
       "      <td>96</td>\n",
       "      <td>13</td>\n",
       "      <td>89</td>\n",
       "      <td>102</td>\n",
       "      <td>0.431373</td>\n",
       "      <td>0.127451</td>\n",
       "      <td>0.872549</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>325</th>\n",
       "      <td>1664</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>326</th>\n",
       "      <td>1670</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>327</th>\n",
       "      <td>1682</td>\n",
       "      <td>15</td>\n",
       "      <td>19</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>1691</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>329</th>\n",
       "      <td>1697</td>\n",
       "      <td>42</td>\n",
       "      <td>16</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>330 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     teacher_id  total_call_before_pt  total_call_after_pt  \\\n",
       "0             9                   128                  152   \n",
       "1            16                     9                   19   \n",
       "2            36                     3                    2   \n",
       "3            37                    41                   29   \n",
       "4            39                   134                  162   \n",
       "..          ...                   ...                  ...   \n",
       "325        1664                     6                    5   \n",
       "326        1670                     6                    4   \n",
       "327        1682                    15                   19   \n",
       "328        1691                     5                    7   \n",
       "329        1697                    42                   16   \n",
       "\n",
       "     call_before_pt_yes  call_after_pt_yes  not_held  held_at_1  held_at_2  \\\n",
       "0                    55                 80        18         61         30   \n",
       "1                     5                  9         2          8          2   \n",
       "2                     2                  2         0          2          0   \n",
       "3                    24                 13        15         10          3   \n",
       "4                    82                 65        44         54         24   \n",
       "..                  ...                ...       ...        ...        ...   \n",
       "325                   5                  5         0          5          0   \n",
       "326                   4                  4         2          2          0   \n",
       "327                  10                  9         2          8          2   \n",
       "328                   4                  5         1          3          1   \n",
       "329                   9                 10         0         10          1   \n",
       "\n",
       "     held_at_3  held_at_4  total_held  buy_premium_y  buy_premium_n  \\\n",
       "0            7          2         100             12             78   \n",
       "1            0          0          10              1             10   \n",
       "2            0          0           2              0              2   \n",
       "3            0          2          15              2             24   \n",
       "4           13          5          96             13             89   \n",
       "..         ...        ...         ...            ...            ...   \n",
       "325          0          0           5              0              5   \n",
       "326          0          0           2              0              4   \n",
       "327          0          0          10              2              8   \n",
       "328          0          0           4              0              5   \n",
       "329          1          0          12              4              6   \n",
       "\n",
       "     total_student  rate_not_held  rate_bp_y  rate_bp_n  fraud_possibility  \n",
       "0               90       0.200000   0.133333   0.866667                  1  \n",
       "1               11       0.181818   0.090909   0.909091                  0  \n",
       "2                2       0.000000   0.000000   1.000000                  0  \n",
       "3               26       0.576923   0.076923   0.923077                  0  \n",
       "4              102       0.431373   0.127451   0.872549                  1  \n",
       "..             ...            ...        ...        ...                ...  \n",
       "325              5       0.000000   0.000000   1.000000                  0  \n",
       "326              4       0.500000   0.000000   1.000000                  0  \n",
       "327             10       0.200000   0.200000   0.800000                  0  \n",
       "328              5       0.200000   0.000000   1.000000                  0  \n",
       "329             10       0.000000   0.400000   0.600000                  0  \n",
       "\n",
       "[330 rows x 18 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_corr = df.copy()\n",
    "\n",
    "fraud_prob = {'Low': 0, 'Medium': 1, 'High': 2}\n",
    "df_corr['fraud_possibility'] = df_corr['fraud_possibility'].replace(fraud_prob)\n",
    "\n",
    "df_corr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "aa50bfdd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:28.561471Z",
     "iopub.status.busy": "2022-10-07T11:58:28.560648Z",
     "iopub.status.idle": "2022-10-07T11:58:28.587437Z",
     "shell.execute_reply": "2022-10-07T11:58:28.586434Z"
    },
    "papermill": {
     "duration": 0.036806,
     "end_time": "2022-10-07T11:58:28.590071",
     "exception": false,
     "start_time": "2022-10-07T11:58:28.553265",
     "status": "completed"
    },
    "tags": []
   },
   "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>teacher_id</th>\n",
       "      <th>total_call_before_pt</th>\n",
       "      <th>total_call_after_pt</th>\n",
       "      <th>call_before_pt_yes</th>\n",
       "      <th>call_after_pt_yes</th>\n",
       "      <th>not_held</th>\n",
       "      <th>held_at_1</th>\n",
       "      <th>held_at_2</th>\n",
       "      <th>held_at_3</th>\n",
       "      <th>held_at_4</th>\n",
       "      <th>total_held</th>\n",
       "      <th>buy_premium_y</th>\n",
       "      <th>buy_premium_n</th>\n",
       "      <th>total_student</th>\n",
       "      <th>rate_not_held</th>\n",
       "      <th>rate_bp_y</th>\n",
       "      <th>rate_bp_n</th>\n",
       "      <th>fraud_possibility</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>teacher_id</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.118935</td>\n",
       "      <td>-0.169332</td>\n",
       "      <td>-0.095289</td>\n",
       "      <td>-0.124720</td>\n",
       "      <td>-0.072269</td>\n",
       "      <td>-0.133051</td>\n",
       "      <td>-0.203857</td>\n",
       "      <td>-0.241166</td>\n",
       "      <td>-0.247250</td>\n",
       "      <td>-0.166692</td>\n",
       "      <td>-0.158516</td>\n",
       "      <td>-0.095461</td>\n",
       "      <td>-0.104043</td>\n",
       "      <td>-0.038407</td>\n",
       "      <td>-0.047381</td>\n",
       "      <td>0.047381</td>\n",
       "      <td>-0.017290</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_call_before_pt</th>\n",
       "      <td>-0.118935</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.967620</td>\n",
       "      <td>0.987832</td>\n",
       "      <td>0.968058</td>\n",
       "      <td>0.973489</td>\n",
       "      <td>0.967933</td>\n",
       "      <td>0.933459</td>\n",
       "      <td>0.830120</td>\n",
       "      <td>0.791392</td>\n",
       "      <td>0.963378</td>\n",
       "      <td>0.871078</td>\n",
       "      <td>0.984898</td>\n",
       "      <td>0.986302</td>\n",
       "      <td>0.129647</td>\n",
       "      <td>-0.094227</td>\n",
       "      <td>0.094227</td>\n",
       "      <td>0.560902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_call_after_pt</th>\n",
       "      <td>-0.169332</td>\n",
       "      <td>0.967620</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.962369</td>\n",
       "      <td>0.983921</td>\n",
       "      <td>0.935535</td>\n",
       "      <td>0.980759</td>\n",
       "      <td>0.974996</td>\n",
       "      <td>0.901314</td>\n",
       "      <td>0.852744</td>\n",
       "      <td>0.990213</td>\n",
       "      <td>0.913510</td>\n",
       "      <td>0.965101</td>\n",
       "      <td>0.973310</td>\n",
       "      <td>0.075851</td>\n",
       "      <td>-0.064216</td>\n",
       "      <td>0.064216</td>\n",
       "      <td>0.504880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>call_before_pt_yes</th>\n",
       "      <td>-0.095289</td>\n",
       "      <td>0.987832</td>\n",
       "      <td>0.962369</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.971773</td>\n",
       "      <td>0.990754</td>\n",
       "      <td>0.971083</td>\n",
       "      <td>0.918213</td>\n",
       "      <td>0.811907</td>\n",
       "      <td>0.777477</td>\n",
       "      <td>0.959965</td>\n",
       "      <td>0.871791</td>\n",
       "      <td>0.996160</td>\n",
       "      <td>0.996526</td>\n",
       "      <td>0.142561</td>\n",
       "      <td>-0.100689</td>\n",
       "      <td>0.100689</td>\n",
       "      <td>0.571794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>call_after_pt_yes</th>\n",
       "      <td>-0.124720</td>\n",
       "      <td>0.968058</td>\n",
       "      <td>0.983921</td>\n",
       "      <td>0.971773</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.946511</td>\n",
       "      <td>0.995875</td>\n",
       "      <td>0.958557</td>\n",
       "      <td>0.851077</td>\n",
       "      <td>0.803333</td>\n",
       "      <td>0.989969</td>\n",
       "      <td>0.906188</td>\n",
       "      <td>0.979273</td>\n",
       "      <td>0.985239</td>\n",
       "      <td>0.077591</td>\n",
       "      <td>-0.087732</td>\n",
       "      <td>0.087732</td>\n",
       "      <td>0.518615</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>not_held</th>\n",
       "      <td>-0.072269</td>\n",
       "      <td>0.973489</td>\n",
       "      <td>0.935535</td>\n",
       "      <td>0.990754</td>\n",
       "      <td>0.946511</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.939400</td>\n",
       "      <td>0.881751</td>\n",
       "      <td>0.774298</td>\n",
       "      <td>0.745013</td>\n",
       "      <td>0.925929</td>\n",
       "      <td>0.846070</td>\n",
       "      <td>0.988168</td>\n",
       "      <td>0.986396</td>\n",
       "      <td>0.190051</td>\n",
       "      <td>-0.111549</td>\n",
       "      <td>0.111549</td>\n",
       "      <td>0.590638</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>held_at_1</th>\n",
       "      <td>-0.133051</td>\n",
       "      <td>0.967933</td>\n",
       "      <td>0.980759</td>\n",
       "      <td>0.971083</td>\n",
       "      <td>0.995875</td>\n",
       "      <td>0.939400</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.959320</td>\n",
       "      <td>0.854112</td>\n",
       "      <td>0.806340</td>\n",
       "      <td>0.993332</td>\n",
       "      <td>0.904829</td>\n",
       "      <td>0.976419</td>\n",
       "      <td>0.982513</td>\n",
       "      <td>0.061217</td>\n",
       "      <td>-0.082861</td>\n",
       "      <td>0.082861</td>\n",
       "      <td>0.508775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>held_at_2</th>\n",
       "      <td>-0.203857</td>\n",
       "      <td>0.933459</td>\n",
       "      <td>0.974996</td>\n",
       "      <td>0.918213</td>\n",
       "      <td>0.958557</td>\n",
       "      <td>0.881751</td>\n",
       "      <td>0.959320</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.913455</td>\n",
       "      <td>0.847547</td>\n",
       "      <td>0.981836</td>\n",
       "      <td>0.912198</td>\n",
       "      <td>0.922258</td>\n",
       "      <td>0.934576</td>\n",
       "      <td>0.050107</td>\n",
       "      <td>-0.041856</td>\n",
       "      <td>0.041856</td>\n",
       "      <td>0.448346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>held_at_3</th>\n",
       "      <td>-0.241166</td>\n",
       "      <td>0.830120</td>\n",
       "      <td>0.901314</td>\n",
       "      <td>0.811907</td>\n",
       "      <td>0.851077</td>\n",
       "      <td>0.774298</td>\n",
       "      <td>0.854112</td>\n",
       "      <td>0.913455</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.866858</td>\n",
       "      <td>0.901207</td>\n",
       "      <td>0.888775</td>\n",
       "      <td>0.804998</td>\n",
       "      <td>0.826300</td>\n",
       "      <td>0.035934</td>\n",
       "      <td>0.007607</td>\n",
       "      <td>-0.007607</td>\n",
       "      <td>0.349408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>held_at_4</th>\n",
       "      <td>-0.247250</td>\n",
       "      <td>0.791392</td>\n",
       "      <td>0.852744</td>\n",
       "      <td>0.777477</td>\n",
       "      <td>0.803333</td>\n",
       "      <td>0.745013</td>\n",
       "      <td>0.806340</td>\n",
       "      <td>0.847547</td>\n",
       "      <td>0.866858</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.850357</td>\n",
       "      <td>0.864704</td>\n",
       "      <td>0.763716</td>\n",
       "      <td>0.786379</td>\n",
       "      <td>0.047435</td>\n",
       "      <td>0.021469</td>\n",
       "      <td>-0.021469</td>\n",
       "      <td>0.343444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_held</th>\n",
       "      <td>-0.166692</td>\n",
       "      <td>0.963378</td>\n",
       "      <td>0.990213</td>\n",
       "      <td>0.959965</td>\n",
       "      <td>0.989969</td>\n",
       "      <td>0.925929</td>\n",
       "      <td>0.993332</td>\n",
       "      <td>0.981836</td>\n",
       "      <td>0.901207</td>\n",
       "      <td>0.850357</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.925289</td>\n",
       "      <td>0.963097</td>\n",
       "      <td>0.972848</td>\n",
       "      <td>0.057458</td>\n",
       "      <td>-0.062916</td>\n",
       "      <td>0.062916</td>\n",
       "      <td>0.486071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>buy_premium_y</th>\n",
       "      <td>-0.158516</td>\n",
       "      <td>0.871078</td>\n",
       "      <td>0.913510</td>\n",
       "      <td>0.871791</td>\n",
       "      <td>0.906188</td>\n",
       "      <td>0.846070</td>\n",
       "      <td>0.904829</td>\n",
       "      <td>0.912198</td>\n",
       "      <td>0.888775</td>\n",
       "      <td>0.864704</td>\n",
       "      <td>0.925289</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.856753</td>\n",
       "      <td>0.885590</td>\n",
       "      <td>0.064534</td>\n",
       "      <td>0.046358</td>\n",
       "      <td>-0.046358</td>\n",
       "      <td>0.410127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>buy_premium_n</th>\n",
       "      <td>-0.095461</td>\n",
       "      <td>0.984898</td>\n",
       "      <td>0.965101</td>\n",
       "      <td>0.996160</td>\n",
       "      <td>0.979273</td>\n",
       "      <td>0.988168</td>\n",
       "      <td>0.976419</td>\n",
       "      <td>0.922258</td>\n",
       "      <td>0.804998</td>\n",
       "      <td>0.763716</td>\n",
       "      <td>0.963097</td>\n",
       "      <td>0.856753</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.998270</td>\n",
       "      <td>0.137893</td>\n",
       "      <td>-0.116470</td>\n",
       "      <td>0.116470</td>\n",
       "      <td>0.572089</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_student</th>\n",
       "      <td>-0.104043</td>\n",
       "      <td>0.986302</td>\n",
       "      <td>0.973310</td>\n",
       "      <td>0.996526</td>\n",
       "      <td>0.985239</td>\n",
       "      <td>0.986396</td>\n",
       "      <td>0.982513</td>\n",
       "      <td>0.934576</td>\n",
       "      <td>0.826300</td>\n",
       "      <td>0.786379</td>\n",
       "      <td>0.972848</td>\n",
       "      <td>0.885590</td>\n",
       "      <td>0.998270</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.131544</td>\n",
       "      <td>-0.099609</td>\n",
       "      <td>0.099609</td>\n",
       "      <td>0.561979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rate_not_held</th>\n",
       "      <td>-0.038407</td>\n",
       "      <td>0.129647</td>\n",
       "      <td>0.075851</td>\n",
       "      <td>0.142561</td>\n",
       "      <td>0.077591</td>\n",
       "      <td>0.190051</td>\n",
       "      <td>0.061217</td>\n",
       "      <td>0.050107</td>\n",
       "      <td>0.035934</td>\n",
       "      <td>0.047435</td>\n",
       "      <td>0.057458</td>\n",
       "      <td>0.064534</td>\n",
       "      <td>0.137893</td>\n",
       "      <td>0.131544</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.316852</td>\n",
       "      <td>0.316852</td>\n",
       "      <td>0.257958</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rate_bp_y</th>\n",
       "      <td>-0.047381</td>\n",
       "      <td>-0.094227</td>\n",
       "      <td>-0.064216</td>\n",
       "      <td>-0.100689</td>\n",
       "      <td>-0.087732</td>\n",
       "      <td>-0.111549</td>\n",
       "      <td>-0.082861</td>\n",
       "      <td>-0.041856</td>\n",
       "      <td>0.007607</td>\n",
       "      <td>0.021469</td>\n",
       "      <td>-0.062916</td>\n",
       "      <td>0.046358</td>\n",
       "      <td>-0.116470</td>\n",
       "      <td>-0.099609</td>\n",
       "      <td>-0.316852</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-0.168926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rate_bp_n</th>\n",
       "      <td>0.047381</td>\n",
       "      <td>0.094227</td>\n",
       "      <td>0.064216</td>\n",
       "      <td>0.100689</td>\n",
       "      <td>0.087732</td>\n",
       "      <td>0.111549</td>\n",
       "      <td>0.082861</td>\n",
       "      <td>0.041856</td>\n",
       "      <td>-0.007607</td>\n",
       "      <td>-0.021469</td>\n",
       "      <td>0.062916</td>\n",
       "      <td>-0.046358</td>\n",
       "      <td>0.116470</td>\n",
       "      <td>0.099609</td>\n",
       "      <td>0.316852</td>\n",
       "      <td>-1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.168926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fraud_possibility</th>\n",
       "      <td>-0.017290</td>\n",
       "      <td>0.560902</td>\n",
       "      <td>0.504880</td>\n",
       "      <td>0.571794</td>\n",
       "      <td>0.518615</td>\n",
       "      <td>0.590638</td>\n",
       "      <td>0.508775</td>\n",
       "      <td>0.448346</td>\n",
       "      <td>0.349408</td>\n",
       "      <td>0.343444</td>\n",
       "      <td>0.486071</td>\n",
       "      <td>0.410127</td>\n",
       "      <td>0.572089</td>\n",
       "      <td>0.561979</td>\n",
       "      <td>0.257958</td>\n",
       "      <td>-0.168926</td>\n",
       "      <td>0.168926</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      teacher_id  total_call_before_pt  total_call_after_pt  \\\n",
       "teacher_id              1.000000             -0.118935            -0.169332   \n",
       "total_call_before_pt   -0.118935              1.000000             0.967620   \n",
       "total_call_after_pt    -0.169332              0.967620             1.000000   \n",
       "call_before_pt_yes     -0.095289              0.987832             0.962369   \n",
       "call_after_pt_yes      -0.124720              0.968058             0.983921   \n",
       "not_held               -0.072269              0.973489             0.935535   \n",
       "held_at_1              -0.133051              0.967933             0.980759   \n",
       "held_at_2              -0.203857              0.933459             0.974996   \n",
       "held_at_3              -0.241166              0.830120             0.901314   \n",
       "held_at_4              -0.247250              0.791392             0.852744   \n",
       "total_held             -0.166692              0.963378             0.990213   \n",
       "buy_premium_y          -0.158516              0.871078             0.913510   \n",
       "buy_premium_n          -0.095461              0.984898             0.965101   \n",
       "total_student          -0.104043              0.986302             0.973310   \n",
       "rate_not_held          -0.038407              0.129647             0.075851   \n",
       "rate_bp_y              -0.047381             -0.094227            -0.064216   \n",
       "rate_bp_n               0.047381              0.094227             0.064216   \n",
       "fraud_possibility      -0.017290              0.560902             0.504880   \n",
       "\n",
       "                      call_before_pt_yes  call_after_pt_yes  not_held  \\\n",
       "teacher_id                     -0.095289          -0.124720 -0.072269   \n",
       "total_call_before_pt            0.987832           0.968058  0.973489   \n",
       "total_call_after_pt             0.962369           0.983921  0.935535   \n",
       "call_before_pt_yes              1.000000           0.971773  0.990754   \n",
       "call_after_pt_yes               0.971773           1.000000  0.946511   \n",
       "not_held                        0.990754           0.946511  1.000000   \n",
       "held_at_1                       0.971083           0.995875  0.939400   \n",
       "held_at_2                       0.918213           0.958557  0.881751   \n",
       "held_at_3                       0.811907           0.851077  0.774298   \n",
       "held_at_4                       0.777477           0.803333  0.745013   \n",
       "total_held                      0.959965           0.989969  0.925929   \n",
       "buy_premium_y                   0.871791           0.906188  0.846070   \n",
       "buy_premium_n                   0.996160           0.979273  0.988168   \n",
       "total_student                   0.996526           0.985239  0.986396   \n",
       "rate_not_held                   0.142561           0.077591  0.190051   \n",
       "rate_bp_y                      -0.100689          -0.087732 -0.111549   \n",
       "rate_bp_n                       0.100689           0.087732  0.111549   \n",
       "fraud_possibility               0.571794           0.518615  0.590638   \n",
       "\n",
       "                      held_at_1  held_at_2  held_at_3  held_at_4  total_held  \\\n",
       "teacher_id            -0.133051  -0.203857  -0.241166  -0.247250   -0.166692   \n",
       "total_call_before_pt   0.967933   0.933459   0.830120   0.791392    0.963378   \n",
       "total_call_after_pt    0.980759   0.974996   0.901314   0.852744    0.990213   \n",
       "call_before_pt_yes     0.971083   0.918213   0.811907   0.777477    0.959965   \n",
       "call_after_pt_yes      0.995875   0.958557   0.851077   0.803333    0.989969   \n",
       "not_held               0.939400   0.881751   0.774298   0.745013    0.925929   \n",
       "held_at_1              1.000000   0.959320   0.854112   0.806340    0.993332   \n",
       "held_at_2              0.959320   1.000000   0.913455   0.847547    0.981836   \n",
       "held_at_3              0.854112   0.913455   1.000000   0.866858    0.901207   \n",
       "held_at_4              0.806340   0.847547   0.866858   1.000000    0.850357   \n",
       "total_held             0.993332   0.981836   0.901207   0.850357    1.000000   \n",
       "buy_premium_y          0.904829   0.912198   0.888775   0.864704    0.925289   \n",
       "buy_premium_n          0.976419   0.922258   0.804998   0.763716    0.963097   \n",
       "total_student          0.982513   0.934576   0.826300   0.786379    0.972848   \n",
       "rate_not_held          0.061217   0.050107   0.035934   0.047435    0.057458   \n",
       "rate_bp_y             -0.082861  -0.041856   0.007607   0.021469   -0.062916   \n",
       "rate_bp_n              0.082861   0.041856  -0.007607  -0.021469    0.062916   \n",
       "fraud_possibility      0.508775   0.448346   0.349408   0.343444    0.486071   \n",
       "\n",
       "                      buy_premium_y  buy_premium_n  total_student  \\\n",
       "teacher_id                -0.158516      -0.095461      -0.104043   \n",
       "total_call_before_pt       0.871078       0.984898       0.986302   \n",
       "total_call_after_pt        0.913510       0.965101       0.973310   \n",
       "call_before_pt_yes         0.871791       0.996160       0.996526   \n",
       "call_after_pt_yes          0.906188       0.979273       0.985239   \n",
       "not_held                   0.846070       0.988168       0.986396   \n",
       "held_at_1                  0.904829       0.976419       0.982513   \n",
       "held_at_2                  0.912198       0.922258       0.934576   \n",
       "held_at_3                  0.888775       0.804998       0.826300   \n",
       "held_at_4                  0.864704       0.763716       0.786379   \n",
       "total_held                 0.925289       0.963097       0.972848   \n",
       "buy_premium_y              1.000000       0.856753       0.885590   \n",
       "buy_premium_n              0.856753       1.000000       0.998270   \n",
       "total_student              0.885590       0.998270       1.000000   \n",
       "rate_not_held              0.064534       0.137893       0.131544   \n",
       "rate_bp_y                  0.046358      -0.116470      -0.099609   \n",
       "rate_bp_n                 -0.046358       0.116470       0.099609   \n",
       "fraud_possibility          0.410127       0.572089       0.561979   \n",
       "\n",
       "                      rate_not_held  rate_bp_y  rate_bp_n  fraud_possibility  \n",
       "teacher_id                -0.038407  -0.047381   0.047381          -0.017290  \n",
       "total_call_before_pt       0.129647  -0.094227   0.094227           0.560902  \n",
       "total_call_after_pt        0.075851  -0.064216   0.064216           0.504880  \n",
       "call_before_pt_yes         0.142561  -0.100689   0.100689           0.571794  \n",
       "call_after_pt_yes          0.077591  -0.087732   0.087732           0.518615  \n",
       "not_held                   0.190051  -0.111549   0.111549           0.590638  \n",
       "held_at_1                  0.061217  -0.082861   0.082861           0.508775  \n",
       "held_at_2                  0.050107  -0.041856   0.041856           0.448346  \n",
       "held_at_3                  0.035934   0.007607  -0.007607           0.349408  \n",
       "held_at_4                  0.047435   0.021469  -0.021469           0.343444  \n",
       "total_held                 0.057458  -0.062916   0.062916           0.486071  \n",
       "buy_premium_y              0.064534   0.046358  -0.046358           0.410127  \n",
       "buy_premium_n              0.137893  -0.116470   0.116470           0.572089  \n",
       "total_student              0.131544  -0.099609   0.099609           0.561979  \n",
       "rate_not_held              1.000000  -0.316852   0.316852           0.257958  \n",
       "rate_bp_y                 -0.316852   1.000000  -1.000000          -0.168926  \n",
       "rate_bp_n                  0.316852  -1.000000   1.000000           0.168926  \n",
       "fraud_possibility          0.257958  -0.168926   0.168926           1.000000  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_corr.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b0317dec",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:28.606087Z",
     "iopub.status.busy": "2022-10-07T11:58:28.605309Z",
     "iopub.status.idle": "2022-10-07T11:58:29.121105Z",
     "shell.execute_reply": "2022-10-07T11:58:29.119886Z"
    },
    "papermill": {
     "duration": 0.526071,
     "end_time": "2022-10-07T11:58:29.123439",
     "exception": false,
     "start_time": "2022-10-07T11:58:28.597368",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(14, 10))\n",
    "sns.heatmap(df_corr.corr(), linewidths=0.1, center=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "22a95fd5",
   "metadata": {
    "papermill": {
     "duration": 0.008528,
     "end_time": "2022-10-07T11:58:29.140155",
     "exception": false,
     "start_time": "2022-10-07T11:58:29.131627",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Making AI Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "a3b3b5ce",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:29.157380Z",
     "iopub.status.busy": "2022-10-07T11:58:29.157083Z",
     "iopub.status.idle": "2022-10-07T11:58:29.172945Z",
     "shell.execute_reply": "2022-10-07T11:58:29.172025Z"
    },
    "papermill": {
     "duration": 0.027175,
     "end_time": "2022-10-07T11:58:29.175126",
     "exception": false,
     "start_time": "2022-10-07T11:58:29.147951",
     "status": "completed"
    },
    "tags": []
   },
   "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>teacher_id</th>\n",
       "      <th>total_call_before_pt</th>\n",
       "      <th>total_call_after_pt</th>\n",
       "      <th>call_before_pt_yes</th>\n",
       "      <th>call_after_pt_yes</th>\n",
       "      <th>not_held</th>\n",
       "      <th>held_at_1</th>\n",
       "      <th>held_at_2</th>\n",
       "      <th>held_at_3</th>\n",
       "      <th>held_at_4</th>\n",
       "      <th>total_held</th>\n",
       "      <th>buy_premium_y</th>\n",
       "      <th>buy_premium_n</th>\n",
       "      <th>total_student</th>\n",
       "      <th>rate_not_held</th>\n",
       "      <th>rate_bp_y</th>\n",
       "      <th>rate_bp_n</th>\n",
       "      <th>fraud_possibility</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9</td>\n",
       "      <td>128</td>\n",
       "      <td>152</td>\n",
       "      <td>55</td>\n",
       "      <td>80</td>\n",
       "      <td>18</td>\n",
       "      <td>61</td>\n",
       "      <td>30</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>100</td>\n",
       "      <td>12</td>\n",
       "      <td>78</td>\n",
       "      <td>90</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.133333</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>16</td>\n",
       "      <td>9</td>\n",
       "      <td>19</td>\n",
       "      <td>5</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "      <td>0.181818</td>\n",
       "      <td>0.090909</td>\n",
       "      <td>0.909091</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>36</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>37</td>\n",
       "      <td>41</td>\n",
       "      <td>29</td>\n",
       "      <td>24</td>\n",
       "      <td>13</td>\n",
       "      <td>15</td>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>2</td>\n",
       "      <td>24</td>\n",
       "      <td>26</td>\n",
       "      <td>0.576923</td>\n",
       "      <td>0.076923</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>39</td>\n",
       "      <td>134</td>\n",
       "      <td>162</td>\n",
       "      <td>82</td>\n",
       "      <td>65</td>\n",
       "      <td>44</td>\n",
       "      <td>54</td>\n",
       "      <td>24</td>\n",
       "      <td>13</td>\n",
       "      <td>5</td>\n",
       "      <td>96</td>\n",
       "      <td>13</td>\n",
       "      <td>89</td>\n",
       "      <td>102</td>\n",
       "      <td>0.431373</td>\n",
       "      <td>0.127451</td>\n",
       "      <td>0.872549</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   teacher_id  total_call_before_pt  total_call_after_pt  call_before_pt_yes  \\\n",
       "0           9                   128                  152                  55   \n",
       "1          16                     9                   19                   5   \n",
       "2          36                     3                    2                   2   \n",
       "3          37                    41                   29                  24   \n",
       "4          39                   134                  162                  82   \n",
       "\n",
       "   call_after_pt_yes  not_held  held_at_1  held_at_2  held_at_3  held_at_4  \\\n",
       "0                 80        18         61         30          7          2   \n",
       "1                  9         2          8          2          0          0   \n",
       "2                  2         0          2          0          0          0   \n",
       "3                 13        15         10          3          0          2   \n",
       "4                 65        44         54         24         13          5   \n",
       "\n",
       "   total_held  buy_premium_y  buy_premium_n  total_student  rate_not_held  \\\n",
       "0         100             12             78             90       0.200000   \n",
       "1          10              1             10             11       0.181818   \n",
       "2           2              0              2              2       0.000000   \n",
       "3          15              2             24             26       0.576923   \n",
       "4          96             13             89            102       0.431373   \n",
       "\n",
       "   rate_bp_y  rate_bp_n  fraud_possibility  \n",
       "0   0.133333   0.866667                  1  \n",
       "1   0.090909   0.909091                  0  \n",
       "2   0.000000   1.000000                  0  \n",
       "3   0.076923   0.923077                  0  \n",
       "4   0.127451   0.872549                  1  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# make a copy of df_corr and save it to a new dataframe\n",
    "dataframe = df_corr.copy()\n",
    "\n",
    "dataframe.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "548826c4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:29.192554Z",
     "iopub.status.busy": "2022-10-07T11:58:29.192288Z",
     "iopub.status.idle": "2022-10-07T11:58:29.201489Z",
     "shell.execute_reply": "2022-10-07T11:58:29.200014Z"
    },
    "papermill": {
     "duration": 0.020843,
     "end_time": "2022-10-07T11:58:29.204163",
     "exception": false,
     "start_time": "2022-10-07T11:58:29.183320",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "297 train examples\n",
      "17 validation examples\n",
      "16 test examples\n"
     ]
    }
   ],
   "source": [
    "train, val = train_test_split(dataframe, train_size=0.95)\n",
    "train, test = train_test_split(train, train_size=0.95)\n",
    "print(len(train), 'train examples')\n",
    "print(len(val), 'validation examples')\n",
    "print(len(test), 'test examples')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "51068c50",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:29.223947Z",
     "iopub.status.busy": "2022-10-07T11:58:29.222381Z",
     "iopub.status.idle": "2022-10-07T11:58:29.228634Z",
     "shell.execute_reply": "2022-10-07T11:58:29.227791Z"
    },
    "papermill": {
     "duration": 0.017709,
     "end_time": "2022-10-07T11:58:29.230626",
     "exception": false,
     "start_time": "2022-10-07T11:58:29.212917",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# transform dataframe to dataset\n",
    "def df_to_dataset(dataframe, target, shuffle=True, batch_size=32):\n",
    "    dataframe = dataframe.copy()\n",
    "    \n",
    "    # Use Pandas dataframe's pop method to get the list of targets.\n",
    "    labels = dataframe.pop(target)\n",
    "    \n",
    "    # Create a tf.data.Dataset from the dataframe and labels.\n",
    "    ds = tf.data.Dataset.from_tensor_slices((dict(dataframe), labels))\n",
    "    \n",
    "    if shuffle:\n",
    "        # Shuffle dataset.\n",
    "        ds = ds.shuffle(buffer_size=len(dataframe))\n",
    "        \n",
    "    # Batch dataset with specified batch_size parameter.\n",
    "    ds = ds.batch(batch_size)\n",
    "    \n",
    "    return ds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f68ba47f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:29.248465Z",
     "iopub.status.busy": "2022-10-07T11:58:29.247726Z",
     "iopub.status.idle": "2022-10-07T11:58:29.254999Z",
     "shell.execute_reply": "2022-10-07T11:58:29.254157Z"
    },
    "papermill": {
     "duration": 0.018258,
     "end_time": "2022-10-07T11:58:29.257016",
     "exception": false,
     "start_time": "2022-10-07T11:58:29.238758",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# numerical columns\n",
    "def get_normalization_layer(name, dataset):\n",
    "    # Create a Normalization layer for the feature.\n",
    "    normalizer = tf.keras.layers.Normalization(axis=None)\n",
    "\n",
    "    # Prepare a Dataset that only yields the feature.\n",
    "    feature_ds = dataset.map(lambda x, y: x[name])\n",
    "\n",
    "    # Learn the statistics of the data.\n",
    "    normalizer.adapt(feature_ds)\n",
    "\n",
    "    return normalizer\n",
    "\n",
    "# encode categorical columns\n",
    "def get_category_encoding_layer(name, dataset, dtype, max_tokens=None):\n",
    "    # Create a layer that turns strings into integer indices.\n",
    "    if dtype == 'string':\n",
    "          index = tf.keras.layers.StringLookup(max_tokens=max_tokens)\n",
    "    # Otherwise, create a layer that turns integer values into integer indices.\n",
    "    else:\n",
    "          index = tf.keras.layers.IntegerLookup(max_tokens=max_tokens)\n",
    "\n",
    "    # Prepare a `tf.data.Dataset` that only yields the feature.\n",
    "    feature_ds = dataset.map(lambda x, y: x[name])\n",
    "\n",
    "    # Learn the set of possible values and assign them a fixed integer index.\n",
    "    index.adapt(feature_ds)\n",
    "\n",
    "    # Encode the integer indices.\n",
    "    encoder = tf.keras.layers.CategoryEncoding(num_tokens=index.vocabulary_size())\n",
    "\n",
    "    # Apply multi-hot encoding to the indices. The lambda function captures the\n",
    "    # layer, so you can use them, or include them in the Keras Functional model later.\n",
    "    return lambda feature: encoder(index(feature))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b84713f0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:29.275040Z",
     "iopub.status.busy": "2022-10-07T11:58:29.274295Z",
     "iopub.status.idle": "2022-10-07T11:58:31.919841Z",
     "shell.execute_reply": "2022-10-07T11:58:31.918854Z"
    },
    "papermill": {
     "duration": 2.657202,
     "end_time": "2022-10-07T11:58:31.922481",
     "exception": false,
     "start_time": "2022-10-07T11:58:29.265279",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-10-07 11:58:29.347493: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-10-07 11:58:29.444714: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-10-07 11:58:29.445581: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-10-07 11:58:29.448340: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 AVX512F FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2022-10-07 11:58:29.448737: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-10-07 11:58:29.449510: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-10-07 11:58:29.450165: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-10-07 11:58:31.504412: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-10-07 11:58:31.505322: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-10-07 11:58:31.506088: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-10-07 11:58:31.506697: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 15401 MB memory:  -> device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:00:04.0, compute capability: 6.0\n"
     ]
    }
   ],
   "source": [
    "batch_size = 32\n",
    "train_ds = df_to_dataset(train, target='fraud_possibility', batch_size=batch_size)\n",
    "val_ds = df_to_dataset(val, target='fraud_possibility', shuffle=False, batch_size=batch_size)\n",
    "test_ds = df_to_dataset(test, target='fraud_possibility', shuffle=False, batch_size=batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "8cade6c2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:31.942163Z",
     "iopub.status.busy": "2022-10-07T11:58:31.941351Z",
     "iopub.status.idle": "2022-10-07T11:58:35.455458Z",
     "shell.execute_reply": "2022-10-07T11:58:35.454469Z"
    },
    "papermill": {
     "duration": 3.526078,
     "end_time": "2022-10-07T11:58:35.458093",
     "exception": false,
     "start_time": "2022-10-07T11:58:31.932015",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-10-07 11:58:32.810090: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)\n"
     ]
    }
   ],
   "source": [
    "all_inputs = []\n",
    "encoded_features = []\n",
    "\n",
    "# numerical columns\n",
    "numerical_col = ['total_call_before_pt',\n",
    "                 'total_call_after_pt',\n",
    "                 'call_before_pt_yes',\n",
    "                 'call_after_pt_yes',\n",
    "                 'not_held',\n",
    "                 'held_at_1',\n",
    "                 'held_at_2',\n",
    "                 'held_at_3',\n",
    "                 'held_at_4',\n",
    "                 'total_held',\n",
    "                 'buy_premium_y',\n",
    "                 'buy_premium_n',\n",
    "                 'total_student',\n",
    "                 'rate_not_held',\n",
    "                 'rate_bp_y',\n",
    "                 'rate_bp_n'\n",
    "                ]\n",
    "\n",
    "for header in numerical_col:\n",
    "    num_col = tf.keras.Input(shape=(1,), name=header)\n",
    "    normalization_layer = get_normalization_layer(header, train_ds)\n",
    "    encoded_numeric_col = normalization_layer(num_col)\n",
    "    all_inputs.append(num_col)\n",
    "    encoded_features.append(encoded_numeric_col)\n",
    "    \n",
    "# categorical column\n",
    "teacher_col = tf.keras.Input(shape=(1,), name='teacher_id', dtype='int64')\n",
    "encoding_layer = get_category_encoding_layer(name='teacher_id',\n",
    "                                             dataset=train_ds,\n",
    "                                             dtype='int64'\n",
    "                                            )\n",
    "encoded_teacher_col = encoding_layer(teacher_col)\n",
    "all_inputs.append(teacher_col)\n",
    "encoded_features.append(encoded_teacher_col)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "3dd52982",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:35.476937Z",
     "iopub.status.busy": "2022-10-07T11:58:35.476649Z",
     "iopub.status.idle": "2022-10-07T11:58:37.060796Z",
     "shell.execute_reply": "2022-10-07T11:58:37.059494Z"
    },
    "papermill": {
     "duration": 1.597852,
     "end_time": "2022-10-07T11:58:37.064682",
     "exception": false,
     "start_time": "2022-10-07T11:58:35.466830",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "teacher_id (InputLayer)         [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "total_call_before_pt (InputLaye [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "total_call_after_pt (InputLayer [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "call_before_pt_yes (InputLayer) [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "call_after_pt_yes (InputLayer)  [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "not_held (InputLayer)           [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "held_at_1 (InputLayer)          [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "held_at_2 (InputLayer)          [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "held_at_3 (InputLayer)          [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "held_at_4 (InputLayer)          [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "total_held (InputLayer)         [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "buy_premium_y (InputLayer)      [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "buy_premium_n (InputLayer)      [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "total_student (InputLayer)      [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "rate_not_held (InputLayer)      [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "rate_bp_y (InputLayer)          [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "rate_bp_n (InputLayer)          [(None, 1)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "integer_lookup (IntegerLookup)  (None, 1)            0           teacher_id[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "normalization (Normalization)   (None, 1)            3           total_call_before_pt[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "normalization_1 (Normalization) (None, 1)            3           total_call_after_pt[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "normalization_2 (Normalization) (None, 1)            3           call_before_pt_yes[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "normalization_3 (Normalization) (None, 1)            3           call_after_pt_yes[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "normalization_4 (Normalization) (None, 1)            3           not_held[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "normalization_5 (Normalization) (None, 1)            3           held_at_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "normalization_6 (Normalization) (None, 1)            3           held_at_2[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "normalization_7 (Normalization) (None, 1)            3           held_at_3[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "normalization_8 (Normalization) (None, 1)            3           held_at_4[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "normalization_9 (Normalization) (None, 1)            3           total_held[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "normalization_10 (Normalization (None, 1)            3           buy_premium_y[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "normalization_11 (Normalization (None, 1)            3           buy_premium_n[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "normalization_12 (Normalization (None, 1)            3           total_student[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "normalization_13 (Normalization (None, 1)            3           rate_not_held[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "normalization_14 (Normalization (None, 1)            3           rate_bp_y[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "normalization_15 (Normalization (None, 1)            3           rate_bp_n[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "category_encoding (CategoryEnco (None, 298)          0           integer_lookup[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "concatenate (Concatenate)       (None, 314)          0           normalization[0][0]              \n",
      "                                                                 normalization_1[0][0]            \n",
      "                                                                 normalization_2[0][0]            \n",
      "                                                                 normalization_3[0][0]            \n",
      "                                                                 normalization_4[0][0]            \n",
      "                                                                 normalization_5[0][0]            \n",
      "                                                                 normalization_6[0][0]            \n",
      "                                                                 normalization_7[0][0]            \n",
      "                                                                 normalization_8[0][0]            \n",
      "                                                                 normalization_9[0][0]            \n",
      "                                                                 normalization_10[0][0]           \n",
      "                                                                 normalization_11[0][0]           \n",
      "                                                                 normalization_12[0][0]           \n",
      "                                                                 normalization_13[0][0]           \n",
      "                                                                 normalization_14[0][0]           \n",
      "                                                                 normalization_15[0][0]           \n",
      "                                                                 category_encoding[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "dense (Dense)                   (None, 128)          40320       concatenate[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "dense_1 (Dense)                 (None, 64)           8256        dense[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "dense_2 (Dense)                 (None, 32)           2080        dense_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_3 (Dense)                 (None, 16)           528         dense_2[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_4 (Dense)                 (None, 8)            136         dense_3[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_5 (Dense)                 (None, 3)            27          dense_4[0][0]                    \n",
      "==================================================================================================\n",
      "Total params: 51,395\n",
      "Trainable params: 51,347\n",
      "Non-trainable params: 48\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "all_features = tf.keras.layers.concatenate(encoded_features)\n",
    "x = tf.keras.layers.Dense(128, activation=\"relu\")(all_features)\n",
    "x = tf.keras.layers.Dense(64, activation=\"relu\")(x)\n",
    "x = tf.keras.layers.Dense(32, activation=\"relu\")(x)\n",
    "x = tf.keras.layers.Dense(16, activation=\"relu\")(x)\n",
    "x = tf.keras.layers.Dense(8, activation=\"relu\")(x)\n",
    "output = tf.keras.layers.Dense(3, activation='softmax')(x)\n",
    "\n",
    "model = tf.keras.Model(all_inputs, output)\n",
    "\n",
    "model.compile(optimizer='adam',\n",
    "              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
    "              metrics=[\"accuracy\"]\n",
    "             )\n",
    "\n",
    "# Use `rankdir='LR'` to make the graph horizontal.\n",
    "tf.keras.utils.plot_model(model, show_shapes=True, rankdir=\"LR\")\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "0f552e97",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:37.084152Z",
     "iopub.status.busy": "2022-10-07T11:58:37.083806Z",
     "iopub.status.idle": "2022-10-07T11:58:41.289638Z",
     "shell.execute_reply": "2022-10-07T11:58:41.288603Z"
    },
    "papermill": {
     "duration": 4.217905,
     "end_time": "2022-10-07T11:58:41.291789",
     "exception": false,
     "start_time": "2022-10-07T11:58:37.073884",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/keras/backend.py:4907: UserWarning: \"`sparse_categorical_crossentropy` received `from_logits=True`, but the `output` argument was produced by a sigmoid or softmax activation and thus does not represent logits. Was this intended?\"\n",
      "  '\"`sparse_categorical_crossentropy` received `from_logits=True`, but '\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10/10 [==============================] - 2s 39ms/step - loss: 1.0509 - accuracy: 0.6768 - val_loss: 0.9684 - val_accuracy: 0.7647\n",
      "Epoch 2/20\n",
      "10/10 [==============================] - 0s 7ms/step - loss: 0.9017 - accuracy: 0.7239 - val_loss: 0.7645 - val_accuracy: 0.7647\n",
      "Epoch 3/20\n",
      "10/10 [==============================] - 0s 7ms/step - loss: 0.6821 - accuracy: 0.7205 - val_loss: 0.5182 - val_accuracy: 0.7647\n",
      "Epoch 4/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.4974 - accuracy: 0.7205 - val_loss: 0.4334 - val_accuracy: 0.7647\n",
      "Epoch 5/20\n",
      "10/10 [==============================] - 0s 7ms/step - loss: 0.4410 - accuracy: 0.7205 - val_loss: 0.4005 - val_accuracy: 0.7647\n",
      "Epoch 6/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.4029 - accuracy: 0.7205 - val_loss: 0.3793 - val_accuracy: 0.7647\n",
      "Epoch 7/20\n",
      "10/10 [==============================] - 0s 7ms/step - loss: 0.3751 - accuracy: 0.7205 - val_loss: 0.3637 - val_accuracy: 0.7647\n",
      "Epoch 8/20\n",
      "10/10 [==============================] - 0s 7ms/step - loss: 0.3559 - accuracy: 0.7374 - val_loss: 0.3444 - val_accuracy: 0.7647\n",
      "Epoch 9/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.3374 - accuracy: 0.7374 - val_loss: 0.3370 - val_accuracy: 0.7647\n",
      "Epoch 10/20\n",
      "10/10 [==============================] - 0s 7ms/step - loss: 0.3231 - accuracy: 0.7542 - val_loss: 0.3367 - val_accuracy: 0.7647\n",
      "Epoch 11/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.3059 - accuracy: 0.7744 - val_loss: 0.3405 - val_accuracy: 0.7647\n",
      "Epoch 12/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.2935 - accuracy: 0.7744 - val_loss: 0.3356 - val_accuracy: 0.7647\n",
      "Epoch 13/20\n",
      "10/10 [==============================] - 0s 8ms/step - loss: 0.2807 - accuracy: 0.8215 - val_loss: 0.3415 - val_accuracy: 0.8235\n",
      "Epoch 14/20\n",
      "10/10 [==============================] - 0s 7ms/step - loss: 0.2644 - accuracy: 0.8620 - val_loss: 0.3531 - val_accuracy: 0.8235\n",
      "Epoch 15/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.2471 - accuracy: 0.9259 - val_loss: 0.3455 - val_accuracy: 0.8235\n",
      "Epoch 16/20\n",
      "10/10 [==============================] - 0s 7ms/step - loss: 0.2294 - accuracy: 0.9731 - val_loss: 0.3629 - val_accuracy: 0.8235\n",
      "Epoch 17/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.2067 - accuracy: 0.9966 - val_loss: 0.3370 - val_accuracy: 0.8235\n",
      "Epoch 18/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.1801 - accuracy: 0.9865 - val_loss: 0.3517 - val_accuracy: 0.8235\n",
      "Epoch 19/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.1495 - accuracy: 0.9933 - val_loss: 0.3247 - val_accuracy: 0.8235\n",
      "Epoch 20/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.1142 - accuracy: 0.9933 - val_loss: 0.3418 - val_accuracy: 0.8235\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7f944cceccd0>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(train_ds, epochs=20, validation_data=val_ds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "8c8b68dc",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:41.316365Z",
     "iopub.status.busy": "2022-10-07T11:58:41.316067Z",
     "iopub.status.idle": "2022-10-07T11:58:41.342151Z",
     "shell.execute_reply": "2022-10-07T11:58:41.340476Z"
    },
    "papermill": {
     "duration": 0.040555,
     "end_time": "2022-10-07T11:58:41.344237",
     "exception": false,
     "start_time": "2022-10-07T11:58:41.303682",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 13ms/step - loss: 0.5511 - accuracy: 0.8125\n",
      "Accuracy 0.8125\n"
     ]
    }
   ],
   "source": [
    "loss, accuracy = model.evaluate(test_ds)\n",
    "print(\"Accuracy\", accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "769fb860",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:41.367838Z",
     "iopub.status.busy": "2022-10-07T11:58:41.366920Z",
     "iopub.status.idle": "2022-10-07T11:58:41.607876Z",
     "shell.execute_reply": "2022-10-07T11:58:41.606868Z"
    },
    "papermill": {
     "duration": 0.254722,
     "end_time": "2022-10-07T11:58:41.610011",
     "exception": false,
     "start_time": "2022-10-07T11:58:41.355289",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[4.5613346e-01, 4.1514868e-01, 1.2871785e-01],\n",
       "       [9.9984193e-01, 1.5780913e-04, 1.9340276e-07],\n",
       "       [9.9999666e-01, 3.3295155e-06, 2.9866365e-10],\n",
       "       [9.9999940e-01, 6.4459965e-07, 6.3176728e-11],\n",
       "       [9.9984598e-01, 1.5386724e-04, 1.3998800e-07],\n",
       "       [9.9995601e-01, 4.4027212e-05, 1.8714088e-08],\n",
       "       [1.9420138e-01, 8.0526578e-01, 5.3288601e-04],\n",
       "       [9.7086990e-01, 2.9101918e-02, 2.8277645e-05],\n",
       "       [9.9999428e-01, 5.7152288e-06, 2.0268853e-09],\n",
       "       [9.9947256e-01, 5.2623259e-04, 1.1495687e-06],\n",
       "       [9.9999988e-01, 9.9610340e-08, 3.4543718e-12],\n",
       "       [1.0000000e+00, 5.5467866e-11, 3.5835118e-17],\n",
       "       [9.9999559e-01, 4.4003618e-06, 1.5652989e-09],\n",
       "       [3.3389005e-01, 6.2365383e-01, 4.2456128e-02],\n",
       "       [3.2824028e-01, 6.3704669e-01, 3.4713034e-02],\n",
       "       [9.9985850e-01, 1.4138181e-04, 1.1421806e-07]], dtype=float32)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prediction = model.predict(test_ds)\n",
    "prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "6482f2da",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:41.634176Z",
     "iopub.status.busy": "2022-10-07T11:58:41.633363Z",
     "iopub.status.idle": "2022-10-07T11:58:41.648944Z",
     "shell.execute_reply": "2022-10-07T11:58:41.647962Z"
    },
    "papermill": {
     "duration": 0.029557,
     "end_time": "2022-10-07T11:58:41.651072",
     "exception": false,
     "start_time": "2022-10-07T11:58:41.621515",
     "status": "completed"
    },
    "tags": []
   },
   "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>fraud_possibility</th>\n",
       "      <th>predict_result</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>194</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>262</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>124</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>311</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     fraud_possibility  predict_result\n",
       "194                  2               0\n",
       "262                  0               0\n",
       "13                   0               0\n",
       "247                  0               0\n",
       "97                   0               0\n",
       "41                   0               0\n",
       "4                    1               1\n",
       "89                   0               0\n",
       "70                   0               0\n",
       "86                   0               0\n",
       "148                  0               0\n",
       "282                  0               0\n",
       "55                   0               0\n",
       "115                  2               1\n",
       "124                  2               1\n",
       "311                  0               0"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compare the prediction with the test data\n",
    "test_result = pd.DataFrame(columns=['fraud_possibility', 'predict_result'])\n",
    "test_result['fraud_possibility'] = test['fraud_possibility']\n",
    "result = []\n",
    "\n",
    "for predict in prediction:\n",
    "    value = predict.max()\n",
    "    index = np.where(predict == value)\n",
    "    result.append(index[0][0])\n",
    "\n",
    "test_result['predict_result'] = result\n",
    "\n",
    "test_result"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f3b76a3",
   "metadata": {
    "papermill": {
     "duration": 0.010952,
     "end_time": "2022-10-07T11:58:41.673083",
     "exception": false,
     "start_time": "2022-10-07T11:58:41.662131",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## 2nd model by excluding teacher_id"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "bc2aac6a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:41.696324Z",
     "iopub.status.busy": "2022-10-07T11:58:41.696060Z",
     "iopub.status.idle": "2022-10-07T11:58:41.717168Z",
     "shell.execute_reply": "2022-10-07T11:58:41.716292Z"
    },
    "papermill": {
     "duration": 0.035044,
     "end_time": "2022-10-07T11:58:41.719229",
     "exception": false,
     "start_time": "2022-10-07T11:58:41.684185",
     "status": "completed"
    },
    "tags": []
   },
   "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>total_call_before_pt</th>\n",
       "      <th>total_call_after_pt</th>\n",
       "      <th>call_before_pt_yes</th>\n",
       "      <th>call_after_pt_yes</th>\n",
       "      <th>not_held</th>\n",
       "      <th>held_at_1</th>\n",
       "      <th>held_at_2</th>\n",
       "      <th>held_at_3</th>\n",
       "      <th>held_at_4</th>\n",
       "      <th>total_held</th>\n",
       "      <th>buy_premium_y</th>\n",
       "      <th>buy_premium_n</th>\n",
       "      <th>total_student</th>\n",
       "      <th>rate_not_held</th>\n",
       "      <th>rate_bp_y</th>\n",
       "      <th>rate_bp_n</th>\n",
       "      <th>fraud_possibility</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>128</td>\n",
       "      <td>152</td>\n",
       "      <td>55</td>\n",
       "      <td>80</td>\n",
       "      <td>18</td>\n",
       "      <td>61</td>\n",
       "      <td>30</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>100</td>\n",
       "      <td>12</td>\n",
       "      <td>78</td>\n",
       "      <td>90</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.133333</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9</td>\n",
       "      <td>19</td>\n",
       "      <td>5</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "      <td>0.181818</td>\n",
       "      <td>0.090909</td>\n",
       "      <td>0.909091</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>41</td>\n",
       "      <td>29</td>\n",
       "      <td>24</td>\n",
       "      <td>13</td>\n",
       "      <td>15</td>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>2</td>\n",
       "      <td>24</td>\n",
       "      <td>26</td>\n",
       "      <td>0.576923</td>\n",
       "      <td>0.076923</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>134</td>\n",
       "      <td>162</td>\n",
       "      <td>82</td>\n",
       "      <td>65</td>\n",
       "      <td>44</td>\n",
       "      <td>54</td>\n",
       "      <td>24</td>\n",
       "      <td>13</td>\n",
       "      <td>5</td>\n",
       "      <td>96</td>\n",
       "      <td>13</td>\n",
       "      <td>89</td>\n",
       "      <td>102</td>\n",
       "      <td>0.431373</td>\n",
       "      <td>0.127451</td>\n",
       "      <td>0.872549</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>325</th>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>326</th>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>327</th>\n",
       "      <td>15</td>\n",
       "      <td>19</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>329</th>\n",
       "      <td>42</td>\n",
       "      <td>16</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>330 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     total_call_before_pt  total_call_after_pt  call_before_pt_yes  \\\n",
       "0                     128                  152                  55   \n",
       "1                       9                   19                   5   \n",
       "2                       3                    2                   2   \n",
       "3                      41                   29                  24   \n",
       "4                     134                  162                  82   \n",
       "..                    ...                  ...                 ...   \n",
       "325                     6                    5                   5   \n",
       "326                     6                    4                   4   \n",
       "327                    15                   19                  10   \n",
       "328                     5                    7                   4   \n",
       "329                    42                   16                   9   \n",
       "\n",
       "     call_after_pt_yes  not_held  held_at_1  held_at_2  held_at_3  held_at_4  \\\n",
       "0                   80        18         61         30          7          2   \n",
       "1                    9         2          8          2          0          0   \n",
       "2                    2         0          2          0          0          0   \n",
       "3                   13        15         10          3          0          2   \n",
       "4                   65        44         54         24         13          5   \n",
       "..                 ...       ...        ...        ...        ...        ...   \n",
       "325                  5         0          5          0          0          0   \n",
       "326                  4         2          2          0          0          0   \n",
       "327                  9         2          8          2          0          0   \n",
       "328                  5         1          3          1          0          0   \n",
       "329                 10         0         10          1          1          0   \n",
       "\n",
       "     total_held  buy_premium_y  buy_premium_n  total_student  rate_not_held  \\\n",
       "0           100             12             78             90       0.200000   \n",
       "1            10              1             10             11       0.181818   \n",
       "2             2              0              2              2       0.000000   \n",
       "3            15              2             24             26       0.576923   \n",
       "4            96             13             89            102       0.431373   \n",
       "..          ...            ...            ...            ...            ...   \n",
       "325           5              0              5              5       0.000000   \n",
       "326           2              0              4              4       0.500000   \n",
       "327          10              2              8             10       0.200000   \n",
       "328           4              0              5              5       0.200000   \n",
       "329          12              4              6             10       0.000000   \n",
       "\n",
       "     rate_bp_y  rate_bp_n  fraud_possibility  \n",
       "0     0.133333   0.866667                  1  \n",
       "1     0.090909   0.909091                  0  \n",
       "2     0.000000   1.000000                  0  \n",
       "3     0.076923   0.923077                  0  \n",
       "4     0.127451   0.872549                  1  \n",
       "..         ...        ...                ...  \n",
       "325   0.000000   1.000000                  0  \n",
       "326   0.000000   1.000000                  0  \n",
       "327   0.200000   0.800000                  0  \n",
       "328   0.000000   1.000000                  0  \n",
       "329   0.400000   0.600000                  0  \n",
       "\n",
       "[330 rows x 17 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# make a copy of df_corr and save it to a another new dataframe\n",
    "dataframe2 = df_corr.copy()\n",
    "dataframe2 = dataframe2.drop(columns=['teacher_id'])\n",
    "\n",
    "dataframe2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "486e4b26",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:41.743733Z",
     "iopub.status.busy": "2022-10-07T11:58:41.742957Z",
     "iopub.status.idle": "2022-10-07T11:58:41.752111Z",
     "shell.execute_reply": "2022-10-07T11:58:41.751019Z"
    },
    "papermill": {
     "duration": 0.024013,
     "end_time": "2022-10-07T11:58:41.754759",
     "exception": false,
     "start_time": "2022-10-07T11:58:41.730746",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "297 train examples\n",
      "17 validation examples\n",
      "16 test examples\n"
     ]
    }
   ],
   "source": [
    "train2, val2 = train_test_split(dataframe2, train_size=0.95)\n",
    "train2, test2 = train_test_split(train2, train_size=0.95)\n",
    "print(len(train2), 'train examples')\n",
    "print(len(val2), 'validation examples')\n",
    "print(len(test2), 'test examples')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "f440427e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:41.779393Z",
     "iopub.status.busy": "2022-10-07T11:58:41.779126Z",
     "iopub.status.idle": "2022-10-07T11:58:41.829053Z",
     "shell.execute_reply": "2022-10-07T11:58:41.828186Z"
    },
    "papermill": {
     "duration": 0.064527,
     "end_time": "2022-10-07T11:58:41.831124",
     "exception": false,
     "start_time": "2022-10-07T11:58:41.766597",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "batch_size = 32\n",
    "train2_ds = df_to_dataset(train2, target='fraud_possibility', batch_size=batch_size)\n",
    "val2_ds = df_to_dataset(val2, target='fraud_possibility', shuffle=False, batch_size=batch_size)\n",
    "test2_ds = df_to_dataset(test2, target='fraud_possibility', shuffle=False, batch_size=batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "606d215f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:41.856177Z",
     "iopub.status.busy": "2022-10-07T11:58:41.855415Z",
     "iopub.status.idle": "2022-10-07T11:58:44.288621Z",
     "shell.execute_reply": "2022-10-07T11:58:44.287678Z"
    },
    "papermill": {
     "duration": 2.448177,
     "end_time": "2022-10-07T11:58:44.291229",
     "exception": false,
     "start_time": "2022-10-07T11:58:41.843052",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "all_inputs2 = []\n",
    "encoded_features2 = []\n",
    "\n",
    "# numerical columns\n",
    "numerical_col2 = ['total_call_before_pt',\n",
    "                 'total_call_after_pt',\n",
    "                 'call_before_pt_yes',\n",
    "                 'call_after_pt_yes',\n",
    "                 'not_held',\n",
    "                 'held_at_1',\n",
    "                 'held_at_2',\n",
    "                 'held_at_3',\n",
    "                 'held_at_4',\n",
    "                 'total_held',\n",
    "                 'buy_premium_y',\n",
    "                 'buy_premium_n',\n",
    "                 'total_student',\n",
    "                 'rate_not_held',\n",
    "                 'rate_bp_y',\n",
    "                 'rate_bp_n'\n",
    "                ]\n",
    "\n",
    "for header in numerical_col2:\n",
    "    num_col = tf.keras.Input(shape=(1,), name=header)\n",
    "    normalization_layer = get_normalization_layer(header, train2_ds)\n",
    "    encoded_numeric_col = normalization_layer(num_col)\n",
    "    all_inputs2.append(num_col)\n",
    "    encoded_features2.append(encoded_numeric_col)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "2b7b96bd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:44.318043Z",
     "iopub.status.busy": "2022-10-07T11:58:44.316222Z",
     "iopub.status.idle": "2022-10-07T11:58:44.375692Z",
     "shell.execute_reply": "2022-10-07T11:58:44.374837Z"
    },
    "papermill": {
     "duration": 0.074583,
     "end_time": "2022-10-07T11:58:44.377859",
     "exception": false,
     "start_time": "2022-10-07T11:58:44.303276",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "all_features2 = tf.keras.layers.concatenate(encoded_features2)\n",
    "x = tf.keras.layers.Dense(128, activation=\"relu\")(all_features2)\n",
    "x = tf.keras.layers.Dense(64, activation=\"relu\")(x)\n",
    "x = tf.keras.layers.Dense(32, activation=\"relu\")(x)\n",
    "x = tf.keras.layers.Dense(16, activation=\"relu\")(x)\n",
    "x = tf.keras.layers.Dense(8, activation=\"relu\")(x)\n",
    "output = tf.keras.layers.Dense(3, activation='softmax')(x)\n",
    "\n",
    "model2 = tf.keras.Model(all_inputs2, output)\n",
    "\n",
    "model2.compile(optimizer='adam',\n",
    "               loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
    "               metrics=[\"accuracy\"]\n",
    "              )\n",
    "\n",
    "# Use `rankdir='LR'` to make the graph horizontal.\n",
    "# tf.keras.utils.plot_model(model2, show_shapes=True, rankdir=\"LR\")\n",
    "\n",
    "# model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "7af14fda",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:44.402391Z",
     "iopub.status.busy": "2022-10-07T11:58:44.402121Z",
     "iopub.status.idle": "2022-10-07T11:58:46.757041Z",
     "shell.execute_reply": "2022-10-07T11:58:46.756139Z"
    },
    "papermill": {
     "duration": 2.369234,
     "end_time": "2022-10-07T11:58:46.758925",
     "exception": false,
     "start_time": "2022-10-07T11:58:44.389691",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/keras/backend.py:4907: UserWarning: \"`sparse_categorical_crossentropy` received `from_logits=True`, but the `output` argument was produced by a sigmoid or softmax activation and thus does not represent logits. Was this intended?\"\n",
      "  '\"`sparse_categorical_crossentropy` received `from_logits=True`, but '\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10/10 [==============================] - 1s 30ms/step - loss: 1.0413 - accuracy: 0.7273 - val_loss: 0.9884 - val_accuracy: 0.7059\n",
      "Epoch 2/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.8896 - accuracy: 0.8721 - val_loss: 0.8486 - val_accuracy: 0.7059\n",
      "Epoch 3/20\n",
      "10/10 [==============================] - 0s 7ms/step - loss: 0.6819 - accuracy: 0.8754 - val_loss: 0.6889 - val_accuracy: 0.7059\n",
      "Epoch 4/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.4747 - accuracy: 0.8721 - val_loss: 0.5889 - val_accuracy: 0.7059\n",
      "Epoch 5/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.3750 - accuracy: 0.8754 - val_loss: 0.5433 - val_accuracy: 0.7059\n",
      "Epoch 6/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.3356 - accuracy: 0.8822 - val_loss: 0.4784 - val_accuracy: 0.7647\n",
      "Epoch 7/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.3021 - accuracy: 0.8889 - val_loss: 0.4176 - val_accuracy: 0.7647\n",
      "Epoch 8/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.2687 - accuracy: 0.8956 - val_loss: 0.3769 - val_accuracy: 0.7647\n",
      "Epoch 9/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.2435 - accuracy: 0.9024 - val_loss: 0.3739 - val_accuracy: 0.7647\n",
      "Epoch 10/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.2211 - accuracy: 0.9024 - val_loss: 0.3459 - val_accuracy: 0.7647\n",
      "Epoch 11/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.2096 - accuracy: 0.9125 - val_loss: 0.3051 - val_accuracy: 0.8235\n",
      "Epoch 12/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.1948 - accuracy: 0.9192 - val_loss: 0.3105 - val_accuracy: 0.8235\n",
      "Epoch 13/20\n",
      "10/10 [==============================] - 0s 6ms/step - loss: 0.1783 - accuracy: 0.9360 - val_loss: 0.3037 - val_accuracy: 0.8235\n",
      "Epoch 14/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.1605 - accuracy: 0.9360 - val_loss: 0.2890 - val_accuracy: 0.8235\n",
      "Epoch 15/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.1481 - accuracy: 0.9495 - val_loss: 0.2747 - val_accuracy: 0.8235\n",
      "Epoch 16/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.1433 - accuracy: 0.9495 - val_loss: 0.2757 - val_accuracy: 0.7647\n",
      "Epoch 17/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.1352 - accuracy: 0.9495 - val_loss: 0.2743 - val_accuracy: 0.8235\n",
      "Epoch 18/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.1296 - accuracy: 0.9562 - val_loss: 0.2622 - val_accuracy: 0.8824\n",
      "Epoch 19/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.1241 - accuracy: 0.9630 - val_loss: 0.2697 - val_accuracy: 0.8824\n",
      "Epoch 20/20\n",
      "10/10 [==============================] - 0s 5ms/step - loss: 0.1264 - accuracy: 0.9529 - val_loss: 0.2833 - val_accuracy: 0.8824\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7f9428092c10>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model2.fit(train2_ds, epochs=20, validation_data=val2_ds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "26dcf71c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:46.790392Z",
     "iopub.status.busy": "2022-10-07T11:58:46.790099Z",
     "iopub.status.idle": "2022-10-07T11:58:46.815104Z",
     "shell.execute_reply": "2022-10-07T11:58:46.814078Z"
    },
    "papermill": {
     "duration": 0.043431,
     "end_time": "2022-10-07T11:58:46.817271",
     "exception": false,
     "start_time": "2022-10-07T11:58:46.773840",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0971 - accuracy: 1.0000\n",
      "Accuracy 1.0\n"
     ]
    }
   ],
   "source": [
    "loss, accuracy = model2.evaluate(test2_ds)\n",
    "print(\"Accuracy\", accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "8046db60",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:46.846641Z",
     "iopub.status.busy": "2022-10-07T11:58:46.846356Z",
     "iopub.status.idle": "2022-10-07T11:58:47.026479Z",
     "shell.execute_reply": "2022-10-07T11:58:47.025570Z"
    },
    "papermill": {
     "duration": 0.197069,
     "end_time": "2022-10-07T11:58:47.028573",
     "exception": false,
     "start_time": "2022-10-07T11:58:46.831504",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[5.74519560e-02, 8.68263423e-01, 7.42846429e-02],\n",
       "       [6.12986922e-01, 3.76933455e-01, 1.00796437e-02],\n",
       "       [9.99765575e-01, 2.34468462e-04, 1.50167310e-08],\n",
       "       [6.19214261e-04, 9.99164939e-01, 2.15803331e-04],\n",
       "       [1.00000000e+00, 1.38805883e-10, 5.80728702e-22],\n",
       "       [9.99996305e-01, 3.73724197e-06, 2.08106427e-12],\n",
       "       [4.87254513e-03, 9.85990644e-01, 9.13674477e-03],\n",
       "       [9.99975681e-01, 2.42690930e-05, 1.16156709e-10],\n",
       "       [1.86365232e-01, 7.44027257e-01, 6.96075708e-02],\n",
       "       [9.99998808e-01, 1.16499439e-06, 1.63225648e-13],\n",
       "       [9.98640954e-01, 1.35836715e-03, 6.66483459e-07],\n",
       "       [9.99925375e-01, 7.46527003e-05, 1.31034106e-09],\n",
       "       [1.53784137e-02, 3.73343557e-01, 6.11278057e-01],\n",
       "       [8.93639266e-01, 1.02212794e-01, 4.14797850e-03],\n",
       "       [9.95050848e-01, 4.93810419e-03, 1.09940838e-05],\n",
       "       [9.98782575e-01, 1.21694128e-03, 5.23396636e-07]], dtype=float32)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prediction2 = model2.predict(test2_ds)\n",
    "prediction2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "dd1dd305",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-10-07T11:58:47.060004Z",
     "iopub.status.busy": "2022-10-07T11:58:47.059107Z",
     "iopub.status.idle": "2022-10-07T11:58:47.074722Z",
     "shell.execute_reply": "2022-10-07T11:58:47.073668Z"
    },
    "papermill": {
     "duration": 0.033353,
     "end_time": "2022-10-07T11:58:47.076752",
     "exception": false,
     "start_time": "2022-10-07T11:58:47.043399",
     "status": "completed"
    },
    "tags": []
   },
   "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>fraud_possibility</th>\n",
       "      <th>predict_result</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>258</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>203</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>329</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>265</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>278</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>213</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     fraud_possibility  predict_result\n",
       "101                  1               1\n",
       "146                  0               0\n",
       "258                  0               0\n",
       "203                  1               1\n",
       "143                  0               0\n",
       "329                  0               0\n",
       "75                   1               1\n",
       "61                   0               0\n",
       "265                  1               1\n",
       "278                  0               0\n",
       "252                  0               0\n",
       "16                   0               0\n",
       "213                  2               2\n",
       "74                   0               0\n",
       "7                    0               0\n",
       "30                   0               0"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compare the prediction with the test data\n",
    "test2_result = pd.DataFrame(columns=['fraud_possibility', 'predict_result'])\n",
    "test2_result['fraud_possibility'] = test2['fraud_possibility']\n",
    "result2 = []\n",
    "\n",
    "for predict in prediction2:\n",
    "    value = predict.max()\n",
    "    index = np.where(predict == value)\n",
    "    result2.append(index[0][0])\n",
    "\n",
    "test2_result['predict_result'] = result2\n",
    "\n",
    "test2_result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b24fa81e",
   "metadata": {
    "papermill": {
     "duration": 0.014517,
     "end_time": "2022-10-07T11:58:47.106002",
     "exception": false,
     "start_time": "2022-10-07T11:58:47.091485",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.12"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 35.267084,
   "end_time": "2022-10-07T11:58:50.721324",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2022-10-07T11:58:15.454240",
   "version": "2.3.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}