Dementia Prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dementia Prediction\n",
    "The objective is to use an ensemble model which will be able to predict whether a person is likely to have dementia or not given their medical records. Various ensemble model i.e. either bagging, stacking or boosting models will be tested against this dataset to find the one which will be best suited for this problem, so that given the symptoms we can predict the outcome of the patient before hand so that precautionary measures can be taken\n",
    "___\n",
    "\n",
    "* Type of Machine Learning Method\n",
    "    * Supervised Learning\n",
    "        * Classfication\n",
    "            * Bagging -Random Forest\n",
    "                * Accuracy - 85.29%\n",
    "                * F1 Score - 82.14%\n",
    "                * Recall - 85.19%\n",
    "            * Stacking - Decision Trees, Naives Bayes, KNN\n",
    "                * Accuracy - 85.29%\n",
    "                * F1 Score - 80.77%\n",
    "                * Recall - 77.78%\n",
    "            * Boosting -  XGBoost\n",
    "                * Accuracy - 86.76%\n",
    "                * F1 Score - 83.64%\n",
    "                * Recall - 85.19%\n",
    "___\n",
    "* Dataset Used\n",
    "     * Data was gotten from [Kaggle]()\n",
    "___\n",
    "\n",
    "### Steps to Solve Problem\n",
    "* Import Dataset and Libraries\n",
    "* Data Preprocessing\n",
    "    * Dropping Unuseful features\n",
    "    * Train / Test Data split\n",
    "    * Feature Engineering\n",
    "    * Missing Data Imputation\n",
    "    * Feature Scaling\n",
    "* Model Build\n",
    "    * Model Initiation and Fitting\n",
    "    * Test predictions\n",
    "* Model Perfromance\n",
    "    * Recall\n",
    "    * Case Prediction\n",
    "        \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import Dataset and Libraries"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd # for data manipulation\n",
    "import numpy as np # for numerical analysis\n",
    "import joblib\n",
    "\n",
    "# for plotting graphs\n",
    "import seaborn as sns \n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Style settings\n",
    "%matplotlib inline\n",
    "sns.set_style(\"darkgrid\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "dementia = pd.read_csv(\"dementia_dataset.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Subject ID</th>\n",
       "      <th>MRI ID</th>\n",
       "      <th>Group</th>\n",
       "      <th>Visit</th>\n",
       "      <th>MR Delay</th>\n",
       "      <th>M/F</th>\n",
       "      <th>Hand</th>\n",
       "      <th>Age</th>\n",
       "      <th>EDUC</th>\n",
       "      <th>SES</th>\n",
       "      <th>MMSE</th>\n",
       "      <th>CDR</th>\n",
       "      <th>eTIV</th>\n",
       "      <th>nWBV</th>\n",
       "      <th>ASF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>OAS2_0001</td>\n",
       "      <td>OAS2_0001_MR1</td>\n",
       "      <td>Nondemented</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>M</td>\n",
       "      <td>R</td>\n",
       "      <td>87</td>\n",
       "      <td>14</td>\n",
       "      <td>2.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1987</td>\n",
       "      <td>0.696</td>\n",
       "      <td>0.883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>OAS2_0001</td>\n",
       "      <td>OAS2_0001_MR2</td>\n",
       "      <td>Nondemented</td>\n",
       "      <td>2</td>\n",
       "      <td>457</td>\n",
       "      <td>M</td>\n",
       "      <td>R</td>\n",
       "      <td>88</td>\n",
       "      <td>14</td>\n",
       "      <td>2.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2004</td>\n",
       "      <td>0.681</td>\n",
       "      <td>0.876</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>OAS2_0002</td>\n",
       "      <td>OAS2_0002_MR1</td>\n",
       "      <td>Demented</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>M</td>\n",
       "      <td>R</td>\n",
       "      <td>75</td>\n",
       "      <td>12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>1678</td>\n",
       "      <td>0.736</td>\n",
       "      <td>1.046</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>OAS2_0002</td>\n",
       "      <td>OAS2_0002_MR2</td>\n",
       "      <td>Demented</td>\n",
       "      <td>2</td>\n",
       "      <td>560</td>\n",
       "      <td>M</td>\n",
       "      <td>R</td>\n",
       "      <td>76</td>\n",
       "      <td>12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>1738</td>\n",
       "      <td>0.713</td>\n",
       "      <td>1.010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>OAS2_0002</td>\n",
       "      <td>OAS2_0002_MR3</td>\n",
       "      <td>Demented</td>\n",
       "      <td>3</td>\n",
       "      <td>1895</td>\n",
       "      <td>M</td>\n",
       "      <td>R</td>\n",
       "      <td>80</td>\n",
       "      <td>12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>22.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>1698</td>\n",
       "      <td>0.701</td>\n",
       "      <td>1.034</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Subject ID         MRI ID        Group  Visit  MR Delay M/F Hand  Age  EDUC  \\\n",
       "0  OAS2_0001  OAS2_0001_MR1  Nondemented      1         0   M    R   87    14   \n",
       "1  OAS2_0001  OAS2_0001_MR2  Nondemented      2       457   M    R   88    14   \n",
       "2  OAS2_0002  OAS2_0002_MR1     Demented      1         0   M    R   75    12   \n",
       "3  OAS2_0002  OAS2_0002_MR2     Demented      2       560   M    R   76    12   \n",
       "4  OAS2_0002  OAS2_0002_MR3     Demented      3      1895   M    R   80    12   \n",
       "\n",
       "   SES  MMSE  CDR  eTIV   nWBV    ASF  \n",
       "0  2.0  27.0  0.0  1987  0.696  0.883  \n",
       "1  2.0  30.0  0.0  2004  0.681  0.876  \n",
       "2  NaN  23.0  0.5  1678  0.736  1.046  \n",
       "3  NaN  28.0  0.5  1738  0.713  1.010  \n",
       "4  NaN  22.0  0.5  1698  0.701  1.034  "
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check first 5 rows of data\n",
    "dementia.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check basic information of the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 373 entries, 0 to 372\n",
      "Data columns (total 15 columns):\n",
      " #   Column      Non-Null Count  Dtype  \n",
      "---  ------      --------------  -----  \n",
      " 0   Subject ID  373 non-null    object \n",
      " 1   MRI ID      373 non-null    object \n",
      " 2   Group       373 non-null    object \n",
      " 3   Visit       373 non-null    int64  \n",
      " 4   MR Delay    373 non-null    int64  \n",
      " 5   M/F         373 non-null    object \n",
      " 6   Hand        373 non-null    object \n",
      " 7   Age         373 non-null    int64  \n",
      " 8   EDUC        373 non-null    int64  \n",
      " 9   SES         354 non-null    float64\n",
      " 10  MMSE        371 non-null    float64\n",
      " 11  CDR         373 non-null    float64\n",
      " 12  eTIV        373 non-null    int64  \n",
      " 13  nWBV        373 non-null    float64\n",
      " 14  ASF         373 non-null    float64\n",
      "dtypes: float64(5), int64(5), object(5)\n",
      "memory usage: 43.8+ KB\n"
     ]
    }
   ],
   "source": [
    "dementia.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Engineering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Mapping Ordinal values to numerical values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "group = {\n",
    "    \"Nondemented\" : 0,\n",
    "    \"Demented\" : 1,\n",
    "    \"Converted\" : 2\n",
    "}\n",
    "dementia[\"Group\"] = dementia[\"Group\"].map(group)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "gender = {\n",
    "    \"M\" : 1,\n",
    "    \"F\" : 0\n",
    "}\n",
    "dementia[\"M/F\"] = dementia[\"M/F\"].map(gender)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Correlations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=[10, 8])\n",
    "sns.heatmap(dementia.corr(), annot=True, cbar=False, cmap=\"Reds\");\n",
    "# (MR delay-Visit*), (educ-ses), (etiv-asf*) high correlated"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The relationship between eTIV and ASF is almost perfectly correlated negatively, which may be the cause of the high correlation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.scatterplot(data=dementia, y=\"eTIV\", x=\"ASF\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see below whenever the number of visits is 1, the MR Delay equals 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "139    0\n",
       "205    0\n",
       "62     0\n",
       "320    0\n",
       "318    0\n",
       "245    0\n",
       "218    0\n",
       "134    0\n",
       "76     0\n",
       "39     0\n",
       "Name: MR Delay, dtype: int64"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dementia[dementia[\"Visit\"] == 1][\"MR Delay\"].sample(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dropping highly correlated data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [],
   "source": [
    "dementia.drop([\"Visit\", \"ASF\"], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Also whenever the CDR feature equals to 1 or 0 its maps to Demented and Non-demented respectively, this column will be dropped to make the model more intelligent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "329    1\n",
       "173    1\n",
       "88     1\n",
       "317    1\n",
       "90     1\n",
       "238    1\n",
       "188    1\n",
       "226    1\n",
       "186    1\n",
       "97     1\n",
       "Name: Group, dtype: int64"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dementia[dementia[\"CDR\"] == 1][\"Group\"].sample(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All patients are right-handed so this feature isnt useful"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "R    373\n",
       "Name: Hand, dtype: int64"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dementia[\"Hand\"].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data Shuffle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [],
   "source": [
    "dementia = dementia.sample(frac=1, random_state=11).copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dropping unuseful features\n",
    "* Subject ID - id of pateint\n",
    "* MRI ID - id of mri scan for each patient\n",
    "* Hand - all patients are right handed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "dementia.drop([\"Subject ID\", \"MRI ID\", \"Hand\", \"CDR\"], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Focus of pateint is Demented or Not Demented, so converted patients are not needed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "mask = dementia.loc[dementia[\"Group\"] == 2].index\n",
    "dementia.drop(mask, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X = dementia.drop(\"Group\", axis=1) \n",
    "y = dementia[\"Group\"]\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=51)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Rows and instances in train and test set respectively"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((268, 8), (68, 8))"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>MR Delay</th>\n",
       "      <th>M/F</th>\n",
       "      <th>Age</th>\n",
       "      <th>EDUC</th>\n",
       "      <th>SES</th>\n",
       "      <th>MMSE</th>\n",
       "      <th>eTIV</th>\n",
       "      <th>nWBV</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>243</th>\n",
       "      <td>1345</td>\n",
       "      <td>1</td>\n",
       "      <td>76</td>\n",
       "      <td>20</td>\n",
       "      <td>2.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1823</td>\n",
       "      <td>0.739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>224</th>\n",
       "      <td>675</td>\n",
       "      <td>1</td>\n",
       "      <td>87</td>\n",
       "      <td>12</td>\n",
       "      <td>4.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1762</td>\n",
       "      <td>0.718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>314</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>78</td>\n",
       "      <td>18</td>\n",
       "      <td>1.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1243</td>\n",
       "      <td>0.748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>69</td>\n",
       "      <td>13</td>\n",
       "      <td>4.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1359</td>\n",
       "      <td>0.789</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>451</td>\n",
       "      <td>1</td>\n",
       "      <td>68</td>\n",
       "      <td>12</td>\n",
       "      <td>4.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>1438</td>\n",
       "      <td>0.738</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>255</th>\n",
       "      <td>597</td>\n",
       "      <td>0</td>\n",
       "      <td>88</td>\n",
       "      <td>16</td>\n",
       "      <td>3.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1295</td>\n",
       "      <td>0.744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>61</td>\n",
       "      <td>16</td>\n",
       "      <td>3.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1313</td>\n",
       "      <td>0.805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>276</th>\n",
       "      <td>539</td>\n",
       "      <td>0</td>\n",
       "      <td>71</td>\n",
       "      <td>11</td>\n",
       "      <td>4.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1284</td>\n",
       "      <td>0.741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>575</td>\n",
       "      <td>0</td>\n",
       "      <td>85</td>\n",
       "      <td>15</td>\n",
       "      <td>2.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1483</td>\n",
       "      <td>0.748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>576</td>\n",
       "      <td>1</td>\n",
       "      <td>69</td>\n",
       "      <td>12</td>\n",
       "      <td>2.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>1480</td>\n",
       "      <td>0.791</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>268 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     MR Delay  M/F  Age  EDUC  SES  MMSE  eTIV   nWBV\n",
       "243      1345    1   76    20  2.0  30.0  1823  0.739\n",
       "224       675    1   87    12  4.0  30.0  1762  0.718\n",
       "314         0    0   78    18  1.0  30.0  1243  0.748\n",
       "68          0    0   69    13  4.0  30.0  1359  0.789\n",
       "142       451    1   68    12  4.0  29.0  1438  0.738\n",
       "..        ...  ...  ...   ...  ...   ...   ...    ...\n",
       "255       597    0   88    16  3.0  30.0  1295  0.744\n",
       "41          0    0   61    16  3.0  30.0  1313  0.805\n",
       "276       539    0   71    11  4.0  28.0  1284  0.741\n",
       "94        575    0   85    15  2.0  22.0  1483  0.748\n",
       "16        576    1   69    12  2.0  24.0  1480  0.791\n",
       "\n",
       "[268 rows x 8 columns]"
      ]
     },
     "execution_count": 130,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Renaming X_train to demnetia\n",
    "dementia = X_train.copy()\n",
    "dementia"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Extracting numerical features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [],
   "source": [
    "numerical = [col for col in dementia.columns if dementia[col].dtype != \"O\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Missing data Imputation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking features that have missing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MR Delay     0\n",
       "M/F          0\n",
       "Age          0\n",
       "EDUC         0\n",
       "SES         14\n",
       "MMSE         0\n",
       "eTIV         0\n",
       "nWBV         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dementia.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Importing library used for handling missing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.impute import SimpleImputer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using median to replace missing values as the data contains outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "imputer = SimpleImputer(strategy=\"median\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Imputing the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "imputer.fit(dementia[numerical])\n",
    "dementia[numerical] = imputer.transform(dementia[numerical])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking features that have missing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MR Delay    0\n",
       "M/F         0\n",
       "Age         0\n",
       "EDUC        0\n",
       "SES         0\n",
       "MMSE        0\n",
       "eTIV        0\n",
       "nWBV        0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 136,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dementia.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Scaling"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Extracting features that will be scaled"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "numerical_scale = ['MR Delay','Age','EDUC','SES','MMSE','eTIV','nWBV']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Importing library for scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Scaling features to a range of -3 to 3 to boost model performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "scaler.fit(dementia[numerical_scale])\n",
    "dementia[numerical_scale] = scaler.transform(dementia[numerical_scale])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['scaler_joblib']"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joblib.dump(scaler, \"scaler_joblib\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>MR Delay</th>\n",
       "      <th>M/F</th>\n",
       "      <th>Age</th>\n",
       "      <th>EDUC</th>\n",
       "      <th>SES</th>\n",
       "      <th>MMSE</th>\n",
       "      <th>eTIV</th>\n",
       "      <th>nWBV</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>243</th>\n",
       "      <td>1.153824</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.048614</td>\n",
       "      <td>1.984618</td>\n",
       "      <td>-0.537880</td>\n",
       "      <td>0.722419</td>\n",
       "      <td>1.757963</td>\n",
       "      <td>0.221513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>224</th>\n",
       "      <td>0.104165</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.384521</td>\n",
       "      <td>-0.861550</td>\n",
       "      <td>1.322144</td>\n",
       "      <td>0.722419</td>\n",
       "      <td>1.430327</td>\n",
       "      <td>-0.348265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>314</th>\n",
       "      <td>-0.953327</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.211956</td>\n",
       "      <td>1.273076</td>\n",
       "      <td>-1.467892</td>\n",
       "      <td>0.722419</td>\n",
       "      <td>-1.357257</td>\n",
       "      <td>0.465703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>-0.953327</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.960609</td>\n",
       "      <td>-0.505779</td>\n",
       "      <td>1.322144</td>\n",
       "      <td>0.722419</td>\n",
       "      <td>-0.734213</td>\n",
       "      <td>1.578127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>-0.246766</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-1.090894</td>\n",
       "      <td>-0.861550</td>\n",
       "      <td>1.322144</td>\n",
       "      <td>0.468340</td>\n",
       "      <td>-0.309899</td>\n",
       "      <td>0.194381</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>255</th>\n",
       "      <td>-0.018034</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.514806</td>\n",
       "      <td>0.561534</td>\n",
       "      <td>0.392132</td>\n",
       "      <td>0.722419</td>\n",
       "      <td>-1.077961</td>\n",
       "      <td>0.357174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>-0.953327</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-2.002889</td>\n",
       "      <td>0.561534</td>\n",
       "      <td>0.392132</td>\n",
       "      <td>0.722419</td>\n",
       "      <td>-0.981282</td>\n",
       "      <td>2.012244</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>276</th>\n",
       "      <td>-0.108900</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.700039</td>\n",
       "      <td>-1.217321</td>\n",
       "      <td>1.322144</td>\n",
       "      <td>0.214261</td>\n",
       "      <td>-1.137043</td>\n",
       "      <td>0.275777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>-0.052500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.123951</td>\n",
       "      <td>0.205763</td>\n",
       "      <td>-0.537880</td>\n",
       "      <td>-1.310215</td>\n",
       "      <td>-0.068201</td>\n",
       "      <td>0.465703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>-0.050934</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.960609</td>\n",
       "      <td>-0.861550</td>\n",
       "      <td>-0.537880</td>\n",
       "      <td>-0.802056</td>\n",
       "      <td>-0.084314</td>\n",
       "      <td>1.632392</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>268 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     MR Delay  M/F       Age      EDUC       SES      MMSE      eTIV      nWBV\n",
       "243  1.153824  1.0 -0.048614  1.984618 -0.537880  0.722419  1.757963  0.221513\n",
       "224  0.104165  1.0  1.384521 -0.861550  1.322144  0.722419  1.430327 -0.348265\n",
       "314 -0.953327  0.0  0.211956  1.273076 -1.467892  0.722419 -1.357257  0.465703\n",
       "68  -0.953327  0.0 -0.960609 -0.505779  1.322144  0.722419 -0.734213  1.578127\n",
       "142 -0.246766  1.0 -1.090894 -0.861550  1.322144  0.468340 -0.309899  0.194381\n",
       "..        ...  ...       ...       ...       ...       ...       ...       ...\n",
       "255 -0.018034  0.0  1.514806  0.561534  0.392132  0.722419 -1.077961  0.357174\n",
       "41  -0.953327  0.0 -2.002889  0.561534  0.392132  0.722419 -0.981282  2.012244\n",
       "276 -0.108900  0.0 -0.700039 -1.217321  1.322144  0.214261 -1.137043  0.275777\n",
       "94  -0.052500  0.0  1.123951  0.205763 -0.537880 -1.310215 -0.068201  0.465703\n",
       "16  -0.050934  1.0 -0.960609 -0.861550 -0.537880 -0.802056 -0.084314  1.632392\n",
       "\n",
       "[268 rows x 8 columns]"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dementia"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Building"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [],
   "source": [
    "# renaming data to X_train\n",
    "X_train = dementia.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Preprocessing Validation data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test[numerical] = imputer.transform(X_test[numerical])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test[numerical_scale] = scaler.transform(X_test[numerical_scale])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>MR Delay</th>\n",
       "      <th>M/F</th>\n",
       "      <th>Age</th>\n",
       "      <th>EDUC</th>\n",
       "      <th>SES</th>\n",
       "      <th>MMSE</th>\n",
       "      <th>eTIV</th>\n",
       "      <th>nWBV</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>213</th>\n",
       "      <td>0.538128</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.309184</td>\n",
       "      <td>1.273076</td>\n",
       "      <td>-0.537880</td>\n",
       "      <td>0.722419</td>\n",
       "      <td>-0.513999</td>\n",
       "      <td>0.574233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0.154298</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.775376</td>\n",
       "      <td>-2.284634</td>\n",
       "      <td>1.322144</td>\n",
       "      <td>-1.056135</td>\n",
       "      <td>0.925447</td>\n",
       "      <td>-2.301790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>-0.184099</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.439469</td>\n",
       "      <td>-0.861550</td>\n",
       "      <td>1.322144</td>\n",
       "      <td>-0.293898</td>\n",
       "      <td>-0.240075</td>\n",
       "      <td>0.709894</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>0.019566</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.993666</td>\n",
       "      <td>-0.150008</td>\n",
       "      <td>-0.537880</td>\n",
       "      <td>-1.310215</td>\n",
       "      <td>0.291661</td>\n",
       "      <td>-1.786276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>222</th>\n",
       "      <td>-0.224832</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.700039</td>\n",
       "      <td>0.561534</td>\n",
       "      <td>-1.467892</td>\n",
       "      <td>-2.580611</td>\n",
       "      <td>0.356114</td>\n",
       "      <td>-1.243631</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     MR Delay  M/F       Age      EDUC       SES      MMSE      eTIV      nWBV\n",
       "213  0.538128  0.0 -0.309184  1.273076 -0.537880  0.722419 -0.513999  0.574233\n",
       "28   0.154298  1.0  1.775376 -2.284634  1.322144 -1.056135  0.925447 -2.301790\n",
       "87  -0.184099  0.0 -0.439469 -0.861550  1.322144 -0.293898 -0.240075  0.709894\n",
       "328  0.019566  1.0  0.993666 -0.150008 -0.537880 -1.310215  0.291661 -1.786276\n",
       "222 -0.224832  1.0 -0.700039  0.561534 -1.467892 -2.580611  0.356114 -1.243631"
      ]
     },
     "execution_count": 145,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Bagging"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Importing library for bagging .i.e Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [],
   "source": [
    "bagger = RandomForestClassifier(random_state=51, n_jobs=-1, n_estimators=40, max_depth=15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [],
   "source": [
    "bagger.fit(X_train, y_train)\n",
    "predictions = bagger.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Bagging Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score, confusion_matrix, f1_score,  recall_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The accuracy is 85.29 %\n",
      "The f1 score is 82.14 %\n",
      "The recall is 85.19 %\n"
     ]
    }
   ],
   "source": [
    "print(f\"The accuracy is {accuracy_score(y_test, predictions) * 100:.2f} %\")\n",
    "print(f\"The f1 score is {f1_score(y_test, predictions) * 100:.2f} %\") \n",
    "print(f\"The recall is {recall_score(y_test, predictions) * 100:.2f} %\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Confusion Matrix\n",
    "|   | 0  | 1  |\n",
    "|---|----|----|\n",
    "| 0 | TN | FP |\n",
    "| 1 | FN | TP |"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.heatmap(confusion_matrix(y_test, predictions), annot=True, cbar=False);\n",
    "# TN   FP\n",
    "# FN*   TP - Recall"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stacking\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Algorithms used for stacking\n",
    "*   Decision Tree\n",
    "*   Naive Bayes\n",
    "*   KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import StackingClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.neighbors import KNeighborsClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "estimators = [\n",
    "    ('dt', DecisionTreeClassifier(max_depth=4, max_leaf_nodes=4, random_state=51)),\n",
    "    ('nb', GaussianNB()),\n",
    "    ('knn', KNeighborsClassifier(n_neighbors=10))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [],
   "source": [
    "stacker = StackingClassifier(estimators)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "StackingClassifier(estimators=[('dt',\n",
       "                                DecisionTreeClassifier(max_depth=4,\n",
       "                                                       max_leaf_nodes=4,\n",
       "                                                       random_state=51)),\n",
       "                               ('nb', GaussianNB()),\n",
       "                               ('knn', KNeighborsClassifier(n_neighbors=10))])"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stacker.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = stacker.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Stacking Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The accuracy is 85.29 %\n",
      "The f1 score is 80.77 %\n",
      "The recall is 77.78 %\n"
     ]
    }
   ],
   "source": [
    "print(f\"The accuracy is {accuracy_score(y_test, predictions) * 100:.2f} %\")\n",
    "print(f\"The f1 score is {f1_score(y_test, predictions) * 100:.2f} %\") \n",
    "print(f\"The recall is {recall_score(y_test, predictions) * 100:.2f} %\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Confusion Matrix\n",
    "|   | 0  | 1  |\n",
    "|---|----|----|\n",
    "| 0 | TN | FP |\n",
    "| 1 | FN | TP |"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.heatmap(confusion_matrix(y_test, predictions), annot=True, cbar=False);\n",
    "# TN   FP\n",
    "# FN*   TP - Recall"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [],
   "source": [
    "from xgboost import XGBClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [],
   "source": [
    "booster = XGBClassifier(random_state=51, n_estimators=60, max_depth=23, learning_rate=0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBClassifier(base_score=0.5, booster='gbtree', callbacks=None,\n",
       "              colsample_bylevel=1, colsample_bynode=1, colsample_bytree=1,\n",
       "              early_stopping_rounds=None, enable_categorical=False,\n",
       "              eval_metric=None, gamma=0, gpu_id=-1, grow_policy='depthwise',\n",
       "              importance_type=None, interaction_constraints='',\n",
       "              learning_rate=0.5, max_bin=256, max_cat_to_onehot=4,\n",
       "              max_delta_step=0, max_depth=23, max_leaves=0, min_child_weight=1,\n",
       "              missing=nan, monotone_constraints='()', n_estimators=60, n_jobs=0,\n",
       "              num_parallel_tree=1, predictor='auto', random_state=51,\n",
       "              reg_alpha=0, reg_lambda=1, ...)"
      ]
     },
     "execution_count": 161,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "booster.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = booster.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Boosting Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The accuracy is 86.76 %\n",
      "The f1 score is 83.64 %\n",
      "The recall is 85.19 %\n"
     ]
    }
   ],
   "source": [
    "print(f\"The accuracy is {accuracy_score(y_test, predictions) * 100:.2f} %\")\n",
    "print(f\"The f1 score is {f1_score(y_test, predictions) * 100:.2f} %\") \n",
    "print(f\"The recall is {recall_score(y_test, predictions) * 100:.2f} %\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Confusion Matrix\n",
    "|   | 0  | 1  |\n",
    "|---|----|----|\n",
    "| 0 | TN | FP |\n",
    "| 1 | FN | TP |"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.heatmap(confusion_matrix(y_test, predictions), annot=True, cbar=False);\n",
    "# TN   FP\n",
    "# FN*   TP - Recall"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Saving The Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['model_joblib']"
      ]
     },
     "execution_count": 165,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joblib.dump(booster, \"model_joblib\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9.12 ('base')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "397fa5c989aefacd265b089cd9fa90a007575229207dba4eead3a1aa550249df"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}