Term Deposit Marketing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d253b399",
   "metadata": {},
   "source": [
    "# Objective - To predict whether a customer will subscribe for a term deposit"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e047847",
   "metadata": {},
   "source": [
    "Import relevant librares"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "e77365da",
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "import os\n",
    "import random\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import shap\n",
    "import tensorflow\n",
    "from imblearn.over_sampling import RandomOverSampler, SMOTE\n",
    "from imblearn.under_sampling import RandomUnderSampler\n",
    "import seaborn as sns\n",
    "import lazypredict\n",
    "from lazypredict.Supervised import LazyClassifier\n",
    "import matplotlib.pyplot as plt\n",
    "import lightgbm\n",
    "from lightgbm.sklearn import LGBMClassifier\n",
    "from lightgbm.sklearn import LGBMRegressor\n",
    "import xgboost\n",
    "from xgboost.sklearn import XGBClassifier\n",
    "from xgboost.sklearn import XGBRegressor\n",
    "import sklearn\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split, cross_val_score\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.ensemble import RandomForestClassifier, StackingClassifier, VotingClassifier\n",
    "from sklearn.metrics import classification_report, r2_score, mean_squared_error\n",
    "from sklearn.linear_model import LogisticRegression, SGDClassifier, PassiveAggressiveClassifier\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.decomposition import PCA\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bad71068",
   "metadata": {},
   "source": [
    "Import data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "3dadae79",
   "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>age</th>\n",
       "      <th>job</th>\n",
       "      <th>marital</th>\n",
       "      <th>education</th>\n",
       "      <th>default</th>\n",
       "      <th>balance</th>\n",
       "      <th>housing</th>\n",
       "      <th>loan</th>\n",
       "      <th>contact</th>\n",
       "      <th>day</th>\n",
       "      <th>month</th>\n",
       "      <th>duration</th>\n",
       "      <th>campaign</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>58</td>\n",
       "      <td>management</td>\n",
       "      <td>married</td>\n",
       "      <td>tertiary</td>\n",
       "      <td>no</td>\n",
       "      <td>2143</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>unknown</td>\n",
       "      <td>5</td>\n",
       "      <td>may</td>\n",
       "      <td>261</td>\n",
       "      <td>1</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>44</td>\n",
       "      <td>technician</td>\n",
       "      <td>single</td>\n",
       "      <td>secondary</td>\n",
       "      <td>no</td>\n",
       "      <td>29</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>unknown</td>\n",
       "      <td>5</td>\n",
       "      <td>may</td>\n",
       "      <td>151</td>\n",
       "      <td>1</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>33</td>\n",
       "      <td>entrepreneur</td>\n",
       "      <td>married</td>\n",
       "      <td>secondary</td>\n",
       "      <td>no</td>\n",
       "      <td>2</td>\n",
       "      <td>yes</td>\n",
       "      <td>yes</td>\n",
       "      <td>unknown</td>\n",
       "      <td>5</td>\n",
       "      <td>may</td>\n",
       "      <td>76</td>\n",
       "      <td>1</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>47</td>\n",
       "      <td>blue-collar</td>\n",
       "      <td>married</td>\n",
       "      <td>unknown</td>\n",
       "      <td>no</td>\n",
       "      <td>1506</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>unknown</td>\n",
       "      <td>5</td>\n",
       "      <td>may</td>\n",
       "      <td>92</td>\n",
       "      <td>1</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>33</td>\n",
       "      <td>unknown</td>\n",
       "      <td>single</td>\n",
       "      <td>unknown</td>\n",
       "      <td>no</td>\n",
       "      <td>1</td>\n",
       "      <td>no</td>\n",
       "      <td>no</td>\n",
       "      <td>unknown</td>\n",
       "      <td>5</td>\n",
       "      <td>may</td>\n",
       "      <td>198</td>\n",
       "      <td>1</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age           job  marital  education default  balance housing loan  \\\n",
       "0   58    management  married   tertiary      no     2143     yes   no   \n",
       "1   44    technician   single  secondary      no       29     yes   no   \n",
       "2   33  entrepreneur  married  secondary      no        2     yes  yes   \n",
       "3   47   blue-collar  married    unknown      no     1506     yes   no   \n",
       "4   33       unknown   single    unknown      no        1      no   no   \n",
       "\n",
       "   contact  day month  duration  campaign   y  \n",
       "0  unknown    5   may       261         1  no  \n",
       "1  unknown    5   may       151         1  no  \n",
       "2  unknown    5   may        76         1  no  \n",
       "3  unknown    5   may        92         1  no  \n",
       "4  unknown    5   may       198         1  no  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.chdir('F:\\Blue Sky\\EVERYTHING\\DATA SCIENCE\\Apziva\\Term Deposit Marketing')\n",
    "tdm = pd.read_csv('term-deposit-marketing-2020.csv')\n",
    "tdm.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "54fc48d7",
   "metadata": {},
   "source": [
    "## Exploratory Data Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "9c07a1e5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 40000 entries, 0 to 39999\n",
      "Data columns (total 14 columns):\n",
      " #   Column     Non-Null Count  Dtype \n",
      "---  ------     --------------  ----- \n",
      " 0   age        40000 non-null  int64 \n",
      " 1   job        40000 non-null  object\n",
      " 2   marital    40000 non-null  object\n",
      " 3   education  40000 non-null  object\n",
      " 4   default    40000 non-null  object\n",
      " 5   balance    40000 non-null  int64 \n",
      " 6   housing    40000 non-null  object\n",
      " 7   loan       40000 non-null  object\n",
      " 8   contact    40000 non-null  object\n",
      " 9   day        40000 non-null  int64 \n",
      " 10  month      40000 non-null  object\n",
      " 11  duration   40000 non-null  int64 \n",
      " 12  campaign   40000 non-null  int64 \n",
      " 13  y          40000 non-null  object\n",
      "dtypes: int64(5), object(9)\n",
      "memory usage: 4.3+ MB\n"
     ]
    }
   ],
   "source": [
    "tdm.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cd4a3e18",
   "metadata": {},
   "source": [
    "Total rows are 40,000. There are no missing values and hence imputation is not requires"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "c32a856e",
   "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>age</th>\n",
       "      <th>job</th>\n",
       "      <th>marital</th>\n",
       "      <th>education</th>\n",
       "      <th>default</th>\n",
       "      <th>balance</th>\n",
       "      <th>housing</th>\n",
       "      <th>loan</th>\n",
       "      <th>contact</th>\n",
       "      <th>day</th>\n",
       "      <th>month</th>\n",
       "      <th>duration</th>\n",
       "      <th>campaign</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>nan</td>\n",
       "      <td>12</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>nan</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>nan</td>\n",
       "      <td>11</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>nan</td>\n",
       "      <td>blue-collar</td>\n",
       "      <td>married</td>\n",
       "      <td>secondary</td>\n",
       "      <td>no</td>\n",
       "      <td>nan</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>cellular</td>\n",
       "      <td>nan</td>\n",
       "      <td>may</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>nan</td>\n",
       "      <td>9383</td>\n",
       "      <td>24386</td>\n",
       "      <td>20993</td>\n",
       "      <td>39191</td>\n",
       "      <td>nan</td>\n",
       "      <td>24031</td>\n",
       "      <td>33070</td>\n",
       "      <td>24914</td>\n",
       "      <td>nan</td>\n",
       "      <td>13532</td>\n",
       "      <td>nan</td>\n",
       "      <td>nan</td>\n",
       "      <td>37104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>40.54</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1274.28</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>16.02</td>\n",
       "      <td>NaN</td>\n",
       "      <td>254.82</td>\n",
       "      <td>2.88</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>9.64</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2903.77</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.28</td>\n",
       "      <td>NaN</td>\n",
       "      <td>259.37</td>\n",
       "      <td>3.24</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>19.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-8019.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>33.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>54.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>100.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>39.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>407.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>17.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>175.00</td>\n",
       "      <td>2.00</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>48.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1319.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>313.00</td>\n",
       "      <td>3.00</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>95.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>102127.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>31.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4918.00</td>\n",
       "      <td>63.00</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            age          job  marital  education default   balance housing  \\\n",
       "count  40000.00        40000    40000      40000   40000  40000.00   40000   \n",
       "unique      nan           12        3          4       2       nan       2   \n",
       "top         nan  blue-collar  married  secondary      no       nan     yes   \n",
       "freq        nan         9383    24386      20993   39191       nan   24031   \n",
       "mean      40.54          NaN      NaN        NaN     NaN   1274.28     NaN   \n",
       "std        9.64          NaN      NaN        NaN     NaN   2903.77     NaN   \n",
       "min       19.00          NaN      NaN        NaN     NaN  -8019.00     NaN   \n",
       "25%       33.00          NaN      NaN        NaN     NaN     54.00     NaN   \n",
       "50%       39.00          NaN      NaN        NaN     NaN    407.00     NaN   \n",
       "75%       48.00          NaN      NaN        NaN     NaN   1319.00     NaN   \n",
       "max       95.00          NaN      NaN        NaN     NaN 102127.00     NaN   \n",
       "\n",
       "         loan   contact      day  month  duration  campaign      y  \n",
       "count   40000     40000 40000.00  40000  40000.00  40000.00  40000  \n",
       "unique      2         3      nan     11       nan       nan      2  \n",
       "top        no  cellular      nan    may       nan       nan     no  \n",
       "freq    33070     24914      nan  13532       nan       nan  37104  \n",
       "mean      NaN       NaN    16.02    NaN    254.82      2.88    NaN  \n",
       "std       NaN       NaN     8.28    NaN    259.37      3.24    NaN  \n",
       "min       NaN       NaN     1.00    NaN      0.00      1.00    NaN  \n",
       "25%       NaN       NaN     8.00    NaN    100.00      1.00    NaN  \n",
       "50%       NaN       NaN    17.00    NaN    175.00      2.00    NaN  \n",
       "75%       NaN       NaN    21.00    NaN    313.00      3.00    NaN  \n",
       "max       NaN       NaN    31.00    NaN   4918.00     63.00    NaN  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tdm.describe(include = 'all')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0affd0e3",
   "metadata": {},
   "source": [
    "Convert categorical data into numeric using dummy variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "5e8f5b57",
   "metadata": {},
   "outputs": [],
   "source": [
    "tdm_job = pd.get_dummies(tdm['job'], prefix = 'job', prefix_sep = '_', drop_first = True)\n",
    "tdm_marital = pd.get_dummies(tdm['marital'], prefix = 'marital', prefix_sep = '_', drop_first = True)\n",
    "tdm_education = pd.get_dummies(tdm['education'], prefix = 'education', prefix_sep = '_', drop_first = True)\n",
    "tdm_default = pd.get_dummies(tdm['default'], prefix = 'default', prefix_sep = '_', drop_first = True)\n",
    "tdm_housing = pd.get_dummies(tdm['housing'], prefix = 'housing', prefix_sep = '_', drop_first = True)\n",
    "tdm_loan = pd.get_dummies(tdm['loan'], prefix = 'loan', prefix_sep = '_', drop_first = True)\n",
    "tdm_contact = pd.get_dummies(tdm['contact'], prefix = 'contact', prefix_sep = '_', drop_first = True)\n",
    "tdm_month = pd.get_dummies(tdm['month'], prefix = 'month', prefix_sep = '_', drop_first = True)\n",
    "tdm_y = pd.get_dummies(tdm['y'], prefix = 'y', prefix_sep = '_', drop_first = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8a930da6",
   "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>age</th>\n",
       "      <th>balance</th>\n",
       "      <th>day</th>\n",
       "      <th>duration</th>\n",
       "      <th>campaign</th>\n",
       "      <th>job_blue-collar</th>\n",
       "      <th>job_entrepreneur</th>\n",
       "      <th>job_housemaid</th>\n",
       "      <th>job_management</th>\n",
       "      <th>job_retired</th>\n",
       "      <th>...</th>\n",
       "      <th>month_dec</th>\n",
       "      <th>month_feb</th>\n",
       "      <th>month_jan</th>\n",
       "      <th>month_jul</th>\n",
       "      <th>month_jun</th>\n",
       "      <th>month_mar</th>\n",
       "      <th>month_may</th>\n",
       "      <th>month_nov</th>\n",
       "      <th>month_oct</th>\n",
       "      <th>y_yes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>58</td>\n",
       "      <td>2143</td>\n",
       "      <td>5</td>\n",
       "      <td>261</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>44</td>\n",
       "      <td>29</td>\n",
       "      <td>5</td>\n",
       "      <td>151</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>33</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>76</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>47</td>\n",
       "      <td>1506</td>\n",
       "      <td>5</td>\n",
       "      <td>92</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>33</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>198</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 37 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   age  balance  day  duration  campaign  job_blue-collar  job_entrepreneur  \\\n",
       "0   58     2143    5       261         1                0                 0   \n",
       "1   44       29    5       151         1                0                 0   \n",
       "2   33        2    5        76         1                0                 1   \n",
       "3   47     1506    5        92         1                1                 0   \n",
       "4   33        1    5       198         1                0                 0   \n",
       "\n",
       "   job_housemaid  job_management  job_retired  ...  month_dec  month_feb  \\\n",
       "0              0               1            0  ...          0          0   \n",
       "1              0               0            0  ...          0          0   \n",
       "2              0               0            0  ...          0          0   \n",
       "3              0               0            0  ...          0          0   \n",
       "4              0               0            0  ...          0          0   \n",
       "\n",
       "   month_jan  month_jul  month_jun  month_mar  month_may  month_nov  \\\n",
       "0          0          0          0          0          1          0   \n",
       "1          0          0          0          0          1          0   \n",
       "2          0          0          0          0          1          0   \n",
       "3          0          0          0          0          1          0   \n",
       "4          0          0          0          0          1          0   \n",
       "\n",
       "   month_oct  y_yes  \n",
       "0          0      0  \n",
       "1          0      0  \n",
       "2          0      0  \n",
       "3          0      0  \n",
       "4          0      0  \n",
       "\n",
       "[5 rows x 37 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Concat original data frame and dummy columns\n",
    "tdm_dc = tdm.drop(columns = ['job', 'marital', 'education', 'default', 'housing', 'loan', 'contact', 'month', 'y'])\n",
    "tdm_dc = pd.concat([tdm_dc, tdm_job, tdm_marital, tdm_education, tdm_default, tdm_housing, tdm_loan, tdm_contact, tdm_month, tdm_y], axis = 1)\n",
    "tdm_dc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "10bf5ee2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x1296 with 42 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize numerical features using histogramsb\n",
    "tdm_dc.hist(figsize = (18,18), grid = False, xlabelsize=9, ylabelsize=9)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "94738508",
   "metadata": {},
   "source": [
    "Here are some observations from the above graphs\n",
    "- The company targets young blue collar workers primarily having having low or negative balance. Usually blue collar people earn less income and that would explain the negative balance.\n",
    "- Most of the calls are in the summer months and peaking in May."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eaa7a27e",
   "metadata": {},
   "source": [
    "Check class balancing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "83e31f8c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    37104\n",
       "1     2896\n",
       "Name: y_yes, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tdm_dc['y_yes'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c201bf5b",
   "metadata": {},
   "source": [
    "The class is heavily skewed towards 'no' which is not surprising. Only a fraction of people will generally subscribe to a term deposit. We'll need to balance our class to make our analysis robust."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "beaaee94",
   "metadata": {},
   "source": [
    "Set seed for reproducability"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "1e409bc4",
   "metadata": {},
   "outputs": [],
   "source": [
    "seed_value = 42\n",
    "random.seed(seed_value)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "992c3f03",
   "metadata": {},
   "source": [
    "Make dataframes for X and y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "339ae982",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = tdm_dc.drop(columns = ['y_yes'])\n",
    "y = tdm_dc['y_yes']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3ecd45ee",
   "metadata": {},
   "source": [
    "Split data for train and test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "c9c21a2c",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = seed_value)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "83f20c83",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(32000, 36)\n",
      "(8000, 36)\n"
     ]
    }
   ],
   "source": [
    "# Check shape of train and test data\n",
    "print(X_train.shape)           \n",
    "print(X_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9dca3c2",
   "metadata": {},
   "source": [
    "## Class balancing"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "691cda09",
   "metadata": {},
   "source": [
    "We'll try different class balancing techniques like oversampling and undersampling. Oversampling will ensure that we don't lose out on majority of the valuable data which has class value as 'no'. We'll experiment with two types of oversampling methods: RandomOverSampler and SMOTE. SMOTE is a smarter way to oversample the imbalanced data."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d43f6147",
   "metadata": {},
   "source": [
    "### a. Oversampling - RandomOverSampler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "df0f54ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "ros = RandomOverSampler(random_state = seed_value)     # Create an instance of RandomOverSampler\n",
    "X_ros, y_ros = ros.fit_resample(X_train, y_train)      # Resample the training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "47e129c6",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    29680\n",
       "0    29680\n",
       "Name: y_yes, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.value_counts(y_ros)                                 # Get counts after class balancing"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c27f2ba",
   "metadata": {},
   "source": [
    "The class is now balanced as counts for both classes are same."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fbf940c2",
   "metadata": {},
   "source": [
    "Lets start with building models using Lazypredict to get an idea of how the models perform. It is an easy way to get F1 score for a large number of models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "f39010b5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get classifier list from lazypredict.Supervised.CLASSIFIERS. We removed LabelSpreading and LabelPropogation models \n",
    "# as they were consuming too much RAM.\n",
    "classifiers = [('AdaBoostClassifier', sklearn.ensemble._weight_boosting.AdaBoostClassifier),\n",
    " ('BaggingClassifier', sklearn.ensemble._bagging.BaggingClassifier),\n",
    " ('BernoulliNB', sklearn.naive_bayes.BernoulliNB),\n",
    " ('CalibratedClassifierCV', sklearn.calibration.CalibratedClassifierCV),\n",
    " ('CategoricalNB', sklearn.naive_bayes.CategoricalNB),\n",
    " ('DecisionTreeClassifier', sklearn.tree._classes.DecisionTreeClassifier),\n",
    " ('DummyClassifier', sklearn.dummy.DummyClassifier),\n",
    " ('ExtraTreeClassifier', sklearn.tree._classes.ExtraTreeClassifier),\n",
    " ('ExtraTreesClassifier', sklearn.ensemble._forest.ExtraTreesClassifier),\n",
    " ('GaussianNB', sklearn.naive_bayes.GaussianNB),\n",
    " ('KNeighborsClassifier',\n",
    "  sklearn.neighbors._classification.KNeighborsClassifier),\n",
    " ('LinearDiscriminantAnalysis',\n",
    "  sklearn.discriminant_analysis.LinearDiscriminantAnalysis),\n",
    " ('LinearSVC', sklearn.svm._classes.LinearSVC),\n",
    " ('LogisticRegression', sklearn.linear_model._logistic.LogisticRegression),\n",
    " ('NearestCentroid', sklearn.neighbors._nearest_centroid.NearestCentroid),\n",
    " ('NuSVC', sklearn.svm._classes.NuSVC),\n",
    " ('PassiveAggressiveClassifier',\n",
    "  sklearn.linear_model._passive_aggressive.PassiveAggressiveClassifier),\n",
    " ('Perceptron', sklearn.linear_model._perceptron.Perceptron),\n",
    " ('QuadraticDiscriminantAnalysis',\n",
    "  sklearn.discriminant_analysis.QuadraticDiscriminantAnalysis),\n",
    " ('RandomForestClassifier', sklearn.ensemble._forest.RandomForestClassifier),\n",
    " ('RidgeClassifier', sklearn.linear_model._ridge.RidgeClassifier),\n",
    " ('RidgeClassifierCV', sklearn.linear_model._ridge.RidgeClassifierCV),\n",
    " ('SGDClassifier', sklearn.linear_model._stochastic_gradient.SGDClassifier),\n",
    " ('SVC', sklearn.svm._classes.SVC),\n",
    " ('StackingClassifier', sklearn.ensemble._stacking.StackingClassifier),\n",
    " ('XGBClassifier', xgboost.sklearn.XGBClassifier),\n",
    " ('LGBMClassifier', lightgbm.sklearn.LGBMClassifier)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "4cc8be53",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Assign new classifier list to lazypredict.Supervised.CLASSIFIERS\n",
    "lazypredict.Supervised.CLASSIFIERS = classifiers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "cbbf09a5",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  4%|███                                                                                | 1/27 [00:05<02:27,  5.65s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'AdaBoostClassifier', 'Accuracy': 0.875375, 'Balanced Accuracy': 0.8663942768199233, 'ROC AUC': 0.8663942768199233, 'F1 Score': 0.8977935964293196, 'Time taken': 5.653932332992554}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  7%|██████▏                                                                            | 2/27 [00:11<02:18,  5.55s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'BaggingClassifier', 'Accuracy': 0.931625, 'Balanced Accuracy': 0.7141403256704981, 'ROC AUC': 0.714140325670498, 'F1 Score': 0.9294160743814535, 'Time taken': 5.415879726409912}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 11%|█████████▏                                                                         | 3/27 [00:11<01:18,  3.26s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'BernoulliNB', 'Accuracy': 0.834375, 'Balanced Accuracy': 0.8394995210727969, 'ROC AUC': 0.839499521072797, 'F1 Score': 0.8687584231920917, 'Time taken': 0.5230364799499512}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 15%|████████████▎                                                                      | 4/27 [01:37<13:41, 35.70s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'CalibratedClassifierCV', 'Accuracy': 0.871, 'Balanced Accuracy': 0.8720441211685823, 'ROC AUC': 0.8720441211685823, 'F1 Score': 0.8949476879554895, 'Time taken': 85.41352319717407}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 22%|██████████████████▍                                                                | 6/27 [01:38<05:23, 15.38s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'DecisionTreeClassifier', 'Accuracy': 0.91825, 'Balanced Accuracy': 0.6701014727011494, 'ROC AUC': 0.6701014727011494, 'F1 Score': 0.9161686298675834, 'Time taken': 0.7887752056121826}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 26%|█████████████████████▌                                                             | 7/27 [01:38<03:28, 10.44s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'DummyClassifier', 'Accuracy': 0.496, 'Balanced Accuracy': 0.5034497725095786, 'ROC AUC': 0.5034497725095786, 'F1 Score': 0.6083372494034699, 'Time taken': 0.2634124755859375}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30%|████████████████████████▌                                                          | 8/27 [01:38<02:17,  7.22s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'ExtraTreeClassifier', 'Accuracy': 0.906125, 'Balanced Accuracy': 0.6299389367816092, 'ROC AUC': 0.6299389367816093, 'F1 Score': 0.9042766709348545, 'Time taken': 0.3239121437072754}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 33%|███████████████████████████▋                                                       | 9/27 [01:48<02:24,  8.00s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'ExtraTreesClassifier', 'Accuracy': 0.937375, 'Balanced Accuracy': 0.6315627993295019, 'ROC AUC': 0.6315627993295019, 'F1 Score': 0.9252266089328309, 'Time taken': 9.729934692382812}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 37%|██████████████████████████████▎                                                   | 10/27 [01:48<01:36,  5.65s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'GaussianNB', 'Accuracy': 0.137875, 'Balanced Accuracy': 0.5314894636015326, 'ROC AUC': 0.5314894636015326, 'F1 Score': 0.1342567774462265, 'Time taken': 0.3698537349700928}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 41%|█████████████████████████████████▍                                                | 11/27 [03:03<07:10, 26.89s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'KNeighborsClassifier', 'Accuracy': 0.891125, 'Balanced Accuracy': 0.7443651221264367, 'ROC AUC': 0.7443651221264367, 'F1 Score': 0.9031751047386654, 'Time taken': 75.06040406227112}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 44%|████████████████████████████████████▍                                             | 12/27 [03:04<04:45, 19.05s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'LinearDiscriminantAnalysis', 'Accuracy': 0.882125, 'Balanced Accuracy': 0.860422653256705, 'ROC AUC': 0.860422653256705, 'F1 Score': 0.902272925284836, 'Time taken': 1.1093039512634277}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 48%|███████████████████████████████████████▍                                          | 13/27 [03:38<05:26, 23.29s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'LinearSVC', 'Accuracy': 0.873375, 'Balanced Accuracy': 0.871722341954023, 'ROC AUC': 0.8717223419540229, 'F1 Score': 0.8965840245353274, 'Time taken': 33.06059956550598}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 52%|██████████████████████████████████████████▌                                       | 14/27 [03:38<03:35, 16.55s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'LogisticRegression', 'Accuracy': 0.872, 'Balanced Accuracy': 0.8709815014367817, 'ROC AUC': 0.8709815014367815, 'F1 Score': 0.8956062194759948, 'Time taken': 0.9826481342315674}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 56%|█████████████████████████████████████████████▌                                    | 15/27 [03:39<02:20, 11.68s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'NearestCentroid', 'Accuracy': 0.825875, 'Balanced Accuracy': 0.8012901101532567, 'ROC AUC': 0.8012901101532567, 'F1 Score': 0.8618273163440782, 'Time taken': 0.36299943923950195}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 59%|████████████████████████████████████████████████                                 | 16/27 [10:11<23:09, 126.33s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'NuSVC', 'Accuracy': 0.87275, 'Balanced Accuracy': 0.8697841834291188, 'ROC AUC': 0.8697841834291188, 'F1 Score': 0.8960867930696307, 'Time taken': 392.59716415405273}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 63%|███████████████████████████████████████████████████▋                              | 17/27 [10:12<14:44, 88.49s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'PassiveAggressiveClassifier', 'Accuracy': 0.814125, 'Balanced Accuracy': 0.7645324473180077, 'ROC AUC': 0.7645324473180077, 'F1 Score': 0.8527380567087142, 'Time taken': 0.48105525970458984}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 67%|██████████████████████████████████████████████████████▋                           | 18/27 [10:12<09:18, 62.01s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'Perceptron', 'Accuracy': 0.8285, 'Balanced Accuracy': 0.8387362308429118, 'ROC AUC': 0.8387362308429119, 'F1 Score': 0.8647104337631889, 'Time taken': 0.35372066497802734}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70%|█████████████████████████████████████████████████████████▋                        | 19/27 [10:13<05:48, 43.54s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'QuadraticDiscriminantAnalysis', 'Accuracy': 0.075875, 'Balanced Accuracy': 0.5012871168582376, 'ROC AUC': 0.5012871168582375, 'F1 Score': 0.017656742639225512, 'Time taken': 0.5168085098266602}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 74%|████████████████████████████████████████████████████████████▋                     | 20/27 [10:22<03:51, 33.13s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'RandomForestClassifier', 'Accuracy': 0.933625, 'Balanced Accuracy': 0.7032073156130269, 'ROC AUC': 0.7032073156130267, 'F1 Score': 0.9299983226893938, 'Time taken': 8.870417833328247}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 78%|███████████████████████████████████████████████████████████████▊                  | 21/27 [10:22<02:19, 23.31s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'RidgeClassifier', 'Accuracy': 0.882125, 'Balanced Accuracy': 0.860422653256705, 'ROC AUC': 0.860422653256705, 'F1 Score': 0.902272925284836, 'Time taken': 0.41939520835876465}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 81%|██████████████████████████████████████████████████████████████████▊               | 22/27 [10:23<01:22, 16.50s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'RidgeClassifierCV', 'Accuracy': 0.882125, 'Balanced Accuracy': 0.860422653256705, 'ROC AUC': 0.860422653256705, 'F1 Score': 0.902272925284836, 'Time taken': 0.6018354892730713}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 85%|█████████████████████████████████████████████████████████████████████▊            | 23/27 [10:24<00:47, 11.81s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'SGDClassifier', 'Accuracy': 0.852875, 'Balanced Accuracy': 0.8654813218390804, 'ROC AUC': 0.8654813218390804, 'F1 Score': 0.8822243706346601, 'Time taken': 0.8816499710083008}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 89%|████████████████████████████████████████████████████████████████████████▉         | 24/27 [13:38<03:19, 66.54s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'SVC', 'Accuracy': 0.88475, 'Balanced Accuracy': 0.8634383979885057, 'ROC AUC': 0.8634383979885056, 'F1 Score': 0.9042085975633192, 'Time taken': 194.1988170146942}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 96%|██████████████████████████████████████████████████████████████████████████████▉   | 26/27 [13:48<00:38, 38.25s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'XGBClassifier', 'Accuracy': 0.90675, 'Balanced Accuracy': 0.8504699473180077, 'ROC AUC': 0.8504699473180077, 'F1 Score': 0.9192042065251823, 'Time taken': 10.494904279708862}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 27/27 [13:50<00:00, 30.76s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'LGBMClassifier', 'Accuracy': 0.88775, 'Balanced Accuracy': 0.8938802083333333, 'ROC AUC': 0.8938802083333333, 'F1 Score': 0.907349608401734, 'Time taken': 1.8665947914123535}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "lzy_o = LazyClassifier(verbose=1, ignore_warnings=True, custom_metric=None)      # Create an instance of LazyClassifier\n",
    "models_o, predictions_o = lzy_o.fit(X_ros, X_test, y_ros, y_test)                # Fit model with training data and make predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "57b79cbc",
   "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>Accuracy</th>\n",
       "      <th>Balanced Accuracy</th>\n",
       "      <th>ROC AUC</th>\n",
       "      <th>F1 Score</th>\n",
       "      <th>Time Taken</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>LGBMClassifier</th>\n",
       "      <td>0.89</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.91</td>\n",
       "      <td>1.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CalibratedClassifierCV</th>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.89</td>\n",
       "      <td>85.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LinearSVC</th>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.90</td>\n",
       "      <td>33.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LogisticRegression</th>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NuSVC</th>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.90</td>\n",
       "      <td>392.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AdaBoostClassifier</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.90</td>\n",
       "      <td>5.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SGDClassifier</th>\n",
       "      <td>0.85</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVC</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.90</td>\n",
       "      <td>194.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RidgeClassifier</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LinearDiscriminantAnalysis</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.90</td>\n",
       "      <td>1.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RidgeClassifierCV</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>XGBClassifier</th>\n",
       "      <td>0.91</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.92</td>\n",
       "      <td>10.49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BernoulliNB</th>\n",
       "      <td>0.83</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Perceptron</th>\n",
       "      <td>0.83</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NearestCentroid</th>\n",
       "      <td>0.83</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PassiveAggressiveClassifier</th>\n",
       "      <td>0.81</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KNeighborsClassifier</th>\n",
       "      <td>0.89</td>\n",
       "      <td>0.74</td>\n",
       "      <td>0.74</td>\n",
       "      <td>0.90</td>\n",
       "      <td>75.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BaggingClassifier</th>\n",
       "      <td>0.93</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.93</td>\n",
       "      <td>5.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RandomForestClassifier</th>\n",
       "      <td>0.93</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.93</td>\n",
       "      <td>8.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DecisionTreeClassifier</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.92</td>\n",
       "      <td>0.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ExtraTreesClassifier</th>\n",
       "      <td>0.94</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.93</td>\n",
       "      <td>9.73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ExtraTreeClassifier</th>\n",
       "      <td>0.91</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GaussianNB</th>\n",
       "      <td>0.14</td>\n",
       "      <td>0.53</td>\n",
       "      <td>0.53</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DummyClassifier</th>\n",
       "      <td>0.50</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>QuadraticDiscriminantAnalysis</th>\n",
       "      <td>0.08</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.52</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               Accuracy  Balanced Accuracy  ROC AUC  F1 Score  \\\n",
       "Model                                                                           \n",
       "LGBMClassifier                     0.89               0.89     0.89      0.91   \n",
       "CalibratedClassifierCV             0.87               0.87     0.87      0.89   \n",
       "LinearSVC                          0.87               0.87     0.87      0.90   \n",
       "LogisticRegression                 0.87               0.87     0.87      0.90   \n",
       "NuSVC                              0.87               0.87     0.87      0.90   \n",
       "AdaBoostClassifier                 0.88               0.87     0.87      0.90   \n",
       "SGDClassifier                      0.85               0.87     0.87      0.88   \n",
       "SVC                                0.88               0.86     0.86      0.90   \n",
       "RidgeClassifier                    0.88               0.86     0.86      0.90   \n",
       "LinearDiscriminantAnalysis         0.88               0.86     0.86      0.90   \n",
       "RidgeClassifierCV                  0.88               0.86     0.86      0.90   \n",
       "XGBClassifier                      0.91               0.85     0.85      0.92   \n",
       "BernoulliNB                        0.83               0.84     0.84      0.87   \n",
       "Perceptron                         0.83               0.84     0.84      0.86   \n",
       "NearestCentroid                    0.83               0.80     0.80      0.86   \n",
       "PassiveAggressiveClassifier        0.81               0.76     0.76      0.85   \n",
       "KNeighborsClassifier               0.89               0.74     0.74      0.90   \n",
       "BaggingClassifier                  0.93               0.71     0.71      0.93   \n",
       "RandomForestClassifier             0.93               0.70     0.70      0.93   \n",
       "DecisionTreeClassifier             0.92               0.67     0.67      0.92   \n",
       "ExtraTreesClassifier               0.94               0.63     0.63      0.93   \n",
       "ExtraTreeClassifier                0.91               0.63     0.63      0.90   \n",
       "GaussianNB                         0.14               0.53     0.53      0.13   \n",
       "DummyClassifier                    0.50               0.50     0.50      0.61   \n",
       "QuadraticDiscriminantAnalysis      0.08               0.50     0.50      0.02   \n",
       "\n",
       "                               Time Taken  \n",
       "Model                                      \n",
       "LGBMClassifier                       1.87  \n",
       "CalibratedClassifierCV              85.41  \n",
       "LinearSVC                           33.06  \n",
       "LogisticRegression                   0.98  \n",
       "NuSVC                              392.60  \n",
       "AdaBoostClassifier                   5.65  \n",
       "SGDClassifier                        0.88  \n",
       "SVC                                194.20  \n",
       "RidgeClassifier                      0.42  \n",
       "LinearDiscriminantAnalysis           1.11  \n",
       "RidgeClassifierCV                    0.60  \n",
       "XGBClassifier                       10.49  \n",
       "BernoulliNB                          0.52  \n",
       "Perceptron                           0.35  \n",
       "NearestCentroid                      0.36  \n",
       "PassiveAggressiveClassifier          0.48  \n",
       "KNeighborsClassifier                75.06  \n",
       "BaggingClassifier                    5.42  \n",
       "RandomForestClassifier               8.87  \n",
       "DecisionTreeClassifier               0.79  \n",
       "ExtraTreesClassifier                 9.73  \n",
       "ExtraTreeClassifier                  0.32  \n",
       "GaussianNB                           0.37  \n",
       "DummyClassifier                      0.26  \n",
       "QuadraticDiscriminantAnalysis        0.52  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "models_o"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a171b2e",
   "metadata": {},
   "source": [
    "### b. Oversampling - SMOTE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "6083ab10",
   "metadata": {},
   "outputs": [],
   "source": [
    "smo = SMOTE(random_state = seed_value)                      # Create an instance of SMOTE\n",
    "X_smo, y_smo = smo.fit_resample(X_train, y_train)           # Resample the training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "18b6ab1d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    29680\n",
       "0    29680\n",
       "Name: y_yes, dtype: int64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.value_counts(y_smo)                                      # Get counts after class balancing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "7a8f4009",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  4%|███                                                                                | 1/27 [00:03<01:39,  3.82s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'AdaBoostClassifier', 'Accuracy': 0.91675, 'Balanced Accuracy': 0.7325490900383143, 'ROC AUC': 0.7325490900383141, 'F1 Score': 0.9200674930895457, 'Time taken': 3.8238563537597656}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  7%|██████▏                                                                            | 2/27 [00:08<01:46,  4.26s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'BaggingClassifier', 'Accuracy': 0.925375, 'Balanced Accuracy': 0.7003636853448276, 'ROC AUC': 0.7003636853448275, 'F1 Score': 0.9237743505635494, 'Time taken': 4.557257413864136}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 11%|█████████▏                                                                         | 3/27 [00:08<00:58,  2.43s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'BernoulliNB', 'Accuracy': 0.917625, 'Balanced Accuracy': 0.6673626077586207, 'ROC AUC': 0.6673626077586207, 'F1 Score': 0.9154906423749236, 'Time taken': 0.25110650062561035}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 19%|███████████████▎                                                                   | 5/27 [00:57<04:50, 13.19s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'CalibratedClassifierCV', 'Accuracy': 0.926375, 'Balanced Accuracy': 0.6920947078544062, 'ROC AUC': 0.6920947078544061, 'F1 Score': 0.9237905942469424, 'Time taken': 48.2983980178833}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 26%|█████████████████████▌                                                             | 7/27 [00:58<02:01,  6.09s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'DecisionTreeClassifier', 'Accuracy': 0.909125, 'Balanced Accuracy': 0.7100245450191571, 'ROC AUC': 0.7100245450191571, 'F1 Score': 0.9131869103955793, 'Time taken': 0.7427699565887451}\n",
      "{'Model': 'DummyClassifier', 'Accuracy': 0.496, 'Balanced Accuracy': 0.5034497725095786, 'ROC AUC': 0.5034497725095786, 'F1 Score': 0.6083372494034699, 'Time taken': 0.1921992301940918}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30%|████████████████████████▌                                                          | 8/27 [00:58<01:20,  4.25s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'ExtraTreeClassifier', 'Accuracy': 0.897, 'Balanced Accuracy': 0.676267660440613, 'ROC AUC': 0.6762676604406129, 'F1 Score': 0.9025329970812435, 'Time taken': 0.2740786075592041}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 33%|███████████████████████████▋                                                       | 9/27 [01:06<01:37,  5.39s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'ExtraTreesClassifier', 'Accuracy': 0.929625, 'Balanced Accuracy': 0.6586147030651341, 'ROC AUC': 0.6586147030651341, 'F1 Score': 0.9229112631563812, 'Time taken': 7.910831928253174}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 37%|██████████████████████████████▎                                                   | 10/27 [01:06<01:04,  3.81s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'GaussianNB', 'Accuracy': 0.599625, 'Balanced Accuracy': 0.5905097581417624, 'ROC AUC': 0.5905097581417625, 'F1 Score': 0.6953604702262187, 'Time taken': 0.26180148124694824}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 41%|█████████████████████████████████▍                                                | 11/27 [02:04<05:26, 20.40s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'KNeighborsClassifier', 'Accuracy': 0.919, 'Balanced Accuracy': 0.6472850814176245, 'ROC AUC': 0.6472850814176245, 'F1 Score': 0.9144502121209916, 'Time taken': 58.01464223861694}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 44%|████████████████████████████████████▍                                             | 12/27 [02:05<03:36, 14.40s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'LinearDiscriminantAnalysis', 'Accuracy': 0.9175, 'Balanced Accuracy': 0.6184521671455938, 'ROC AUC': 0.6184521671455939, 'F1 Score': 0.9102598102763545, 'Time taken': 0.6934077739715576}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 48%|███████████████████████████████████████▍                                          | 13/27 [02:19<03:19, 14.22s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'LinearSVC', 'Accuracy': 0.92325, 'Balanced Accuracy': 0.7144321719348659, 'ROC AUC': 0.7144321719348659, 'F1 Score': 0.9233724445886587, 'Time taken': 13.803860425949097}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 52%|██████████████████████████████████████████▌                                       | 14/27 [02:19<02:10, 10.06s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'LogisticRegression', 'Accuracy': 0.92375, 'Balanced Accuracy': 0.7187051005747127, 'ROC AUC': 0.7187051005747127, 'F1 Score': 0.9240524039430787, 'Time taken': 0.44660019874572754}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 56%|█████████████████████████████████████████████▌                                    | 15/27 [02:19<01:25,  7.10s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'NearestCentroid', 'Accuracy': 0.913875, 'Balanced Accuracy': 0.678153436302682, 'ROC AUC': 0.678153436302682, 'F1 Score': 0.9139094478044366, 'Time taken': 0.2301192283630371}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 59%|████████████████████████████████████████████████                                 | 16/27 [08:47<22:17, 121.63s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'NuSVC', 'Accuracy': 0.920375, 'Balanced Accuracy': 0.5455355004789272, 'ROC AUC': 0.5455355004789272, 'F1 Score': 0.9009409348842907, 'Time taken': 387.6055061817169}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 63%|███████████████████████████████████████████████████▋                              | 17/27 [08:47<14:11, 85.16s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'PassiveAggressiveClassifier', 'Accuracy': 0.9035, 'Balanced Accuracy': 0.7278122006704981, 'ROC AUC': 0.727812200670498, 'F1 Score': 0.9105634913478836, 'Time taken': 0.3389303684234619}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 67%|██████████████████████████████████████████████████████▋                           | 18/27 [08:48<08:57, 59.70s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'Perceptron', 'Accuracy': 0.8795, 'Balanced Accuracy': 0.6828529094827587, 'ROC AUC': 0.6828529094827586, 'F1 Score': 0.891655570312119, 'Time taken': 0.4242842197418213}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70%|█████████████████████████████████████████████████████████▋                        | 19/27 [08:48<05:35, 41.91s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'QuadraticDiscriminantAnalysis', 'Accuracy': 0.35725, 'Balanced Accuracy': 0.5391897150383143, 'ROC AUC': 0.5391897150383143, 'F1 Score': 0.46081851473189206, 'Time taken': 0.4842848777770996}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 74%|████████████████████████████████████████████████████████████▋                     | 20/27 [08:57<03:44, 32.14s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'RandomForestClassifier', 'Accuracy': 0.9315, 'Balanced Accuracy': 0.7100694444444444, 'ROC AUC': 0.7100694444444444, 'F1 Score': 0.9289988532282678, 'Time taken': 9.346187114715576}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 78%|███████████████████████████████████████████████████████████████▊                  | 21/27 [08:58<02:15, 22.57s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'RidgeClassifier', 'Accuracy': 0.9175, 'Balanced Accuracy': 0.6184521671455938, 'ROC AUC': 0.6184521671455939, 'F1 Score': 0.9102598102763545, 'Time taken': 0.28240227699279785}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 81%|██████████████████████████████████████████████████████████████████▊               | 22/27 [08:58<01:20, 16.00s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'RidgeClassifierCV', 'Accuracy': 0.9175, 'Balanced Accuracy': 0.6184521671455938, 'ROC AUC': 0.6184521671455939, 'F1 Score': 0.9102598102763545, 'Time taken': 0.6681251525878906}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 85%|█████████████████████████████████████████████████████████████████████▊            | 23/27 [08:59<00:45, 11.36s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'SGDClassifier', 'Accuracy': 0.921375, 'Balanced Accuracy': 0.7030127514367817, 'ROC AUC': 0.7030127514367817, 'F1 Score': 0.9211531864794911, 'Time taken': 0.528594970703125}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 89%|████████████████████████████████████████████████████████████████████████▉         | 24/27 [10:04<01:22, 27.54s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'SVC', 'Accuracy': 0.932, 'Balanced Accuracy': 0.7159437859195402, 'ROC AUC': 0.7159437859195403, 'F1 Score': 0.929834658343361, 'Time taken': 65.29525399208069}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 96%|██████████████████████████████████████████████████████████████████████████████▉   | 26/27 [10:12<00:16, 16.72s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'XGBClassifier', 'Accuracy': 0.926125, 'Balanced Accuracy': 0.7432052203065134, 'ROC AUC': 0.7432052203065134, 'F1 Score': 0.9275097240575755, 'Time taken': 8.210760116577148}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 27/27 [10:14<00:00, 22.75s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'LGBMClassifier', 'Accuracy': 0.935, 'Balanced Accuracy': 0.7816166786398467, 'ROC AUC': 0.7816166786398467, 'F1 Score': 0.9365235247465711, 'Time taken': 1.480285406112671}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "lzy_smo = LazyClassifier(verbose=1, ignore_warnings=True, custom_metric=None)      # Create an instance of LazyClassifier\n",
    "models_smo ,predictions_smo = lzy_smo.fit(X_smo, X_test, y_smo, y_test)            # Fit model with training data and make predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "9087ef5a",
   "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>Accuracy</th>\n",
       "      <th>Balanced Accuracy</th>\n",
       "      <th>ROC AUC</th>\n",
       "      <th>F1 Score</th>\n",
       "      <th>Time Taken</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>LGBMClassifier</th>\n",
       "      <td>0.94</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.94</td>\n",
       "      <td>1.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>XGBClassifier</th>\n",
       "      <td>0.93</td>\n",
       "      <td>0.74</td>\n",
       "      <td>0.74</td>\n",
       "      <td>0.93</td>\n",
       "      <td>8.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AdaBoostClassifier</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.92</td>\n",
       "      <td>3.82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PassiveAggressiveClassifier</th>\n",
       "      <td>0.90</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.91</td>\n",
       "      <td>0.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LogisticRegression</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.72</td>\n",
       "      <td>0.72</td>\n",
       "      <td>0.92</td>\n",
       "      <td>0.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVC</th>\n",
       "      <td>0.93</td>\n",
       "      <td>0.72</td>\n",
       "      <td>0.72</td>\n",
       "      <td>0.93</td>\n",
       "      <td>65.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LinearSVC</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.92</td>\n",
       "      <td>13.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RandomForestClassifier</th>\n",
       "      <td>0.93</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.93</td>\n",
       "      <td>9.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DecisionTreeClassifier</th>\n",
       "      <td>0.91</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.91</td>\n",
       "      <td>0.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SGDClassifier</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.92</td>\n",
       "      <td>0.53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BaggingClassifier</th>\n",
       "      <td>0.93</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.92</td>\n",
       "      <td>4.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CalibratedClassifierCV</th>\n",
       "      <td>0.93</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.92</td>\n",
       "      <td>48.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Perceptron</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.68</td>\n",
       "      <td>0.68</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NearestCentroid</th>\n",
       "      <td>0.91</td>\n",
       "      <td>0.68</td>\n",
       "      <td>0.68</td>\n",
       "      <td>0.91</td>\n",
       "      <td>0.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ExtraTreeClassifier</th>\n",
       "      <td>0.90</td>\n",
       "      <td>0.68</td>\n",
       "      <td>0.68</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BernoulliNB</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.92</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ExtraTreesClassifier</th>\n",
       "      <td>0.93</td>\n",
       "      <td>0.66</td>\n",
       "      <td>0.66</td>\n",
       "      <td>0.92</td>\n",
       "      <td>7.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KNeighborsClassifier</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.65</td>\n",
       "      <td>0.65</td>\n",
       "      <td>0.91</td>\n",
       "      <td>58.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LinearDiscriminantAnalysis</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.62</td>\n",
       "      <td>0.62</td>\n",
       "      <td>0.91</td>\n",
       "      <td>0.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RidgeClassifier</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.62</td>\n",
       "      <td>0.62</td>\n",
       "      <td>0.91</td>\n",
       "      <td>0.28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RidgeClassifierCV</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.62</td>\n",
       "      <td>0.62</td>\n",
       "      <td>0.91</td>\n",
       "      <td>0.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GaussianNB</th>\n",
       "      <td>0.60</td>\n",
       "      <td>0.59</td>\n",
       "      <td>0.59</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NuSVC</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.55</td>\n",
       "      <td>0.55</td>\n",
       "      <td>0.90</td>\n",
       "      <td>387.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>QuadraticDiscriminantAnalysis</th>\n",
       "      <td>0.36</td>\n",
       "      <td>0.54</td>\n",
       "      <td>0.54</td>\n",
       "      <td>0.46</td>\n",
       "      <td>0.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DummyClassifier</th>\n",
       "      <td>0.50</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               Accuracy  Balanced Accuracy  ROC AUC  F1 Score  \\\n",
       "Model                                                                           \n",
       "LGBMClassifier                     0.94               0.78     0.78      0.94   \n",
       "XGBClassifier                      0.93               0.74     0.74      0.93   \n",
       "AdaBoostClassifier                 0.92               0.73     0.73      0.92   \n",
       "PassiveAggressiveClassifier        0.90               0.73     0.73      0.91   \n",
       "LogisticRegression                 0.92               0.72     0.72      0.92   \n",
       "SVC                                0.93               0.72     0.72      0.93   \n",
       "LinearSVC                          0.92               0.71     0.71      0.92   \n",
       "RandomForestClassifier             0.93               0.71     0.71      0.93   \n",
       "DecisionTreeClassifier             0.91               0.71     0.71      0.91   \n",
       "SGDClassifier                      0.92               0.70     0.70      0.92   \n",
       "BaggingClassifier                  0.93               0.70     0.70      0.92   \n",
       "CalibratedClassifierCV             0.93               0.69     0.69      0.92   \n",
       "Perceptron                         0.88               0.68     0.68      0.89   \n",
       "NearestCentroid                    0.91               0.68     0.68      0.91   \n",
       "ExtraTreeClassifier                0.90               0.68     0.68      0.90   \n",
       "BernoulliNB                        0.92               0.67     0.67      0.92   \n",
       "ExtraTreesClassifier               0.93               0.66     0.66      0.92   \n",
       "KNeighborsClassifier               0.92               0.65     0.65      0.91   \n",
       "LinearDiscriminantAnalysis         0.92               0.62     0.62      0.91   \n",
       "RidgeClassifier                    0.92               0.62     0.62      0.91   \n",
       "RidgeClassifierCV                  0.92               0.62     0.62      0.91   \n",
       "GaussianNB                         0.60               0.59     0.59      0.70   \n",
       "NuSVC                              0.92               0.55     0.55      0.90   \n",
       "QuadraticDiscriminantAnalysis      0.36               0.54     0.54      0.46   \n",
       "DummyClassifier                    0.50               0.50     0.50      0.61   \n",
       "\n",
       "                               Time Taken  \n",
       "Model                                      \n",
       "LGBMClassifier                       1.48  \n",
       "XGBClassifier                        8.21  \n",
       "AdaBoostClassifier                   3.82  \n",
       "PassiveAggressiveClassifier          0.34  \n",
       "LogisticRegression                   0.45  \n",
       "SVC                                 65.30  \n",
       "LinearSVC                           13.80  \n",
       "RandomForestClassifier               9.35  \n",
       "DecisionTreeClassifier               0.74  \n",
       "SGDClassifier                        0.53  \n",
       "BaggingClassifier                    4.56  \n",
       "CalibratedClassifierCV              48.30  \n",
       "Perceptron                           0.42  \n",
       "NearestCentroid                      0.23  \n",
       "ExtraTreeClassifier                  0.27  \n",
       "BernoulliNB                          0.25  \n",
       "ExtraTreesClassifier                 7.91  \n",
       "KNeighborsClassifier                58.01  \n",
       "LinearDiscriminantAnalysis           0.69  \n",
       "RidgeClassifier                      0.28  \n",
       "RidgeClassifierCV                    0.67  \n",
       "GaussianNB                           0.26  \n",
       "NuSVC                              387.61  \n",
       "QuadraticDiscriminantAnalysis        0.48  \n",
       "DummyClassifier                      0.19  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "models_smo"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "976801e2",
   "metadata": {},
   "source": [
    "### c. Undersampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "9f3b3329",
   "metadata": {},
   "outputs": [],
   "source": [
    "rus = RandomUnderSampler(random_state = seed_value)               # Create an instance of RandomUnderSampler\n",
    "X_rus, y_rus = rus.fit_resample(X_train, y_train)                 # Resample training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "63f2aa54",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    2320\n",
       "0    2320\n",
       "Name: y_yes, dtype: int64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.value_counts(y_rus)                                            # Check value counts after rebalancing"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5db75b64",
   "metadata": {},
   "source": [
    "The classes is now balanced as both classes have equal counts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "2ff1c5d4",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  4%|███                                                                                | 1/27 [00:00<00:13,  1.86it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'AdaBoostClassifier', 'Accuracy': 0.876625, 'Balanced Accuracy': 0.8638649425287357, 'ROC AUC': 0.8638649425287357, 'F1 Score': 0.8985742666235176, 'Time taken': 0.5382215976715088}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  7%|██████▏                                                                            | 2/27 [00:00<00:11,  2.17it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'BaggingClassifier', 'Accuracy': 0.8625, 'Balanced Accuracy': 0.8618594348659003, 'ROC AUC': 0.8618594348659004, 'F1 Score': 0.8887448851260208, 'Time taken': 0.4082486629486084}\n",
      "{'Model': 'BernoulliNB', 'Accuracy': 0.831625, 'Balanced Accuracy': 0.8380178400383143, 'ROC AUC': 0.8380178400383141, 'F1 Score': 0.8668324441125772, 'Time taken': 0.0514833927154541}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 22%|██████████████████▍                                                                | 6/27 [00:03<00:09,  2.11it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'CalibratedClassifierCV', 'Accuracy': 0.87325, 'Balanced Accuracy': 0.8724556992337165, 'ROC AUC': 0.8724556992337164, 'F1 Score': 0.8965213623802597, 'Time taken': 2.112241506576538}\n",
      "{'Model': 'DecisionTreeClassifier', 'Accuracy': 0.81875, 'Balanced Accuracy': 0.8102625119731801, 'ROC AUC': 0.8102625119731801, 'F1 Score': 0.8572486167007612, 'Time taken': 0.08149099349975586}\n",
      "{'Model': 'DummyClassifier', 'Accuracy': 0.496, 'Balanced Accuracy': 0.5034497725095786, 'ROC AUC': 0.5034497725095786, 'F1 Score': 0.6083372494034699, 'Time taken': 0.050662994384765625}\n",
      "{'Model': 'ExtraTreeClassifier', 'Accuracy': 0.764875, 'Balanced Accuracy': 0.7500074832375478, 'ROC AUC': 0.7500074832375478, 'F1 Score': 0.8188073021303063, 'Time taken': 0.04893994331359863}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 33%|███████████████████████████▋                                                       | 9/27 [00:04<00:07,  2.46it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'ExtraTreesClassifier', 'Accuracy': 0.85725, 'Balanced Accuracy': 0.862233596743295, 'ROC AUC': 0.862233596743295, 'F1 Score': 0.8851394953132069, 'Time taken': 0.9254708290100098}\n",
      "{'Model': 'GaussianNB', 'Accuracy': 0.922125, 'Balanced Accuracy': 0.5753038194444444, 'ROC AUC': 0.5753038194444444, 'F1 Score': 0.9071473755149402, 'Time taken': 0.05049300193786621}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 41%|█████████████████████████████████▍                                                | 11/27 [00:07<00:12,  1.28it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'KNeighborsClassifier', 'Accuracy': 0.796625, 'Balanced Accuracy': 0.7591071000957854, 'ROC AUC': 0.7591071000957854, 'F1 Score': 0.8407689191585589, 'Time taken': 3.071946620941162}\n",
      "{'Model': 'LinearDiscriminantAnalysis', 'Accuracy': 0.8815, 'Balanced Accuracy': 0.8608866139846744, 'ROC AUC': 0.8608866139846744, 'F1 Score': 0.901854917921778, 'Time taken': 0.08058929443359375}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 56%|█████████████████████████████████████████████▌                                    | 15/27 [00:08<00:05,  2.13it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'LinearSVC', 'Accuracy': 0.8755, 'Balanced Accuracy': 0.8736679837164751, 'ROC AUC': 0.873667983716475, 'F1 Score': 0.8981234285714286, 'Time taken': 0.6754498481750488}\n",
      "{'Model': 'LogisticRegression', 'Accuracy': 0.873375, 'Balanced Accuracy': 0.8701209291187739, 'ROC AUC': 0.8701209291187739, 'F1 Score': 0.8965312481260139, 'Time taken': 0.08222174644470215}\n",
      "{'Model': 'NearestCentroid', 'Accuracy': 0.829875, 'Balanced Accuracy': 0.8106516403256705, 'ROC AUC': 0.8106516403256705, 'F1 Score': 0.8648304628256862, 'Time taken': 0.04687213897705078}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 67%|██████████████████████████████████████████████████████▋                           | 18/27 [00:10<00:05,  1.77it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'NuSVC', 'Accuracy': 0.862, 'Balanced Accuracy': 0.8663942768199233, 'ROC AUC': 0.8663942768199233, 'F1 Score': 0.8885438757429911, 'Time taken': 2.4586100578308105}\n",
      "{'Model': 'PassiveAggressiveClassifier', 'Accuracy': 0.8095, 'Balanced Accuracy': 0.7812574832375478, 'ROC AUC': 0.7812574832375478, 'F1 Score': 0.8501280859368794, 'Time taken': 0.0518646240234375}\n",
      "{'Model': 'Perceptron', 'Accuracy': 0.837625, 'Balanced Accuracy': 0.8084216355363985, 'ROC AUC': 0.8084216355363983, 'F1 Score': 0.8700165313971313, 'Time taken': 0.05690813064575195}\n",
      "{'Model': 'QuadraticDiscriminantAnalysis', 'Accuracy': 0.9275, 'Balanced Accuracy': 0.5445701628352491, 'ROC AUC': 0.5445701628352492, 'F1 Score': 0.9044924789033201, 'Time taken': 0.08442544937133789}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 81%|██████████████████████████████████████████████████████████████████▊               | 22/27 [00:11<00:01,  2.57it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'RandomForestClassifier', 'Accuracy': 0.86625, 'Balanced Accuracy': 0.8854989822796935, 'ROC AUC': 0.8854989822796935, 'F1 Score': 0.8920733879800169, 'Time taken': 0.9068706035614014}\n",
      "{'Model': 'RidgeClassifier', 'Accuracy': 0.8815, 'Balanced Accuracy': 0.8608866139846744, 'ROC AUC': 0.8608866139846744, 'F1 Score': 0.901854917921778, 'Time taken': 0.05094265937805176}\n",
      "{'Model': 'RidgeClassifierCV', 'Accuracy': 0.88175, 'Balanced Accuracy': 0.8610213122605364, 'ROC AUC': 0.8610213122605364, 'F1 Score': 0.9020335182577225, 'Time taken': 0.07262110710144043}\n",
      "{'Model': 'SGDClassifier', 'Accuracy': 0.85375, 'Balanced Accuracy': 0.858746408045977, 'ROC AUC': 0.858746408045977, 'F1 Score': 0.88262314094665, 'Time taken': 0.09061193466186523}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 89%|████████████████████████████████████████████████████████████████████████▉         | 24/27 [00:13<00:01,  1.72it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'SVC', 'Accuracy': 0.860875, 'Balanced Accuracy': 0.8713930795019158, 'ROC AUC': 0.8713930795019156, 'F1 Score': 0.8879222665364546, 'Time taken': 1.9251954555511475}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 96%|██████████████████████████████████████████████████████████████████████████████▉   | 26/27 [00:14<00:00,  1.92it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'XGBClassifier', 'Accuracy': 0.86825, 'Balanced Accuracy': 0.8929822198275862, 'ROC AUC': 0.8929822198275862, 'F1 Score': 0.8936872438606895, 'Time taken': 0.7720274925231934}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 27/27 [00:15<00:00,  1.78it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Model': 'LGBMClassifier', 'Accuracy': 0.869625, 'Balanced Accuracy': 0.8945237667624522, 'ROC AUC': 0.8945237667624523, 'F1 Score': 0.8946903378526866, 'Time taken': 0.4239811897277832}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "lzy_u = LazyClassifier(verbose = 1, ignore_warnings = True, custom_metric = None)     # Create an instance of LazyClassifier\n",
    "models_u, predictions_u = lzy_u.fit(X_rus, X_test, y_rus, y_test)                     # Fit model with training data and make predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "05eff9a7",
   "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>Accuracy</th>\n",
       "      <th>Balanced Accuracy</th>\n",
       "      <th>ROC AUC</th>\n",
       "      <th>F1 Score</th>\n",
       "      <th>Time Taken</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>LGBMClassifier</th>\n",
       "      <td>0.87</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>XGBClassifier</th>\n",
       "      <td>0.87</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RandomForestClassifier</th>\n",
       "      <td>0.87</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LinearSVC</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CalibratedClassifierCV</th>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.90</td>\n",
       "      <td>2.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVC</th>\n",
       "      <td>0.86</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.89</td>\n",
       "      <td>1.93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LogisticRegression</th>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NuSVC</th>\n",
       "      <td>0.86</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.89</td>\n",
       "      <td>2.46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AdaBoostClassifier</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ExtraTreesClassifier</th>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BaggingClassifier</th>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RidgeClassifierCV</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LinearDiscriminantAnalysis</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RidgeClassifier</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SGDClassifier</th>\n",
       "      <td>0.85</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BernoulliNB</th>\n",
       "      <td>0.83</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NearestCentroid</th>\n",
       "      <td>0.83</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DecisionTreeClassifier</th>\n",
       "      <td>0.82</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Perceptron</th>\n",
       "      <td>0.84</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PassiveAggressiveClassifier</th>\n",
       "      <td>0.81</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KNeighborsClassifier</th>\n",
       "      <td>0.80</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.84</td>\n",
       "      <td>3.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ExtraTreeClassifier</th>\n",
       "      <td>0.76</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GaussianNB</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.58</td>\n",
       "      <td>0.58</td>\n",
       "      <td>0.91</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>QuadraticDiscriminantAnalysis</th>\n",
       "      <td>0.93</td>\n",
       "      <td>0.54</td>\n",
       "      <td>0.54</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DummyClassifier</th>\n",
       "      <td>0.50</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               Accuracy  Balanced Accuracy  ROC AUC  F1 Score  \\\n",
       "Model                                                                           \n",
       "LGBMClassifier                     0.87               0.89     0.89      0.89   \n",
       "XGBClassifier                      0.87               0.89     0.89      0.89   \n",
       "RandomForestClassifier             0.87               0.89     0.89      0.89   \n",
       "LinearSVC                          0.88               0.87     0.87      0.90   \n",
       "CalibratedClassifierCV             0.87               0.87     0.87      0.90   \n",
       "SVC                                0.86               0.87     0.87      0.89   \n",
       "LogisticRegression                 0.87               0.87     0.87      0.90   \n",
       "NuSVC                              0.86               0.87     0.87      0.89   \n",
       "AdaBoostClassifier                 0.88               0.86     0.86      0.90   \n",
       "ExtraTreesClassifier               0.86               0.86     0.86      0.89   \n",
       "BaggingClassifier                  0.86               0.86     0.86      0.89   \n",
       "RidgeClassifierCV                  0.88               0.86     0.86      0.90   \n",
       "LinearDiscriminantAnalysis         0.88               0.86     0.86      0.90   \n",
       "RidgeClassifier                    0.88               0.86     0.86      0.90   \n",
       "SGDClassifier                      0.85               0.86     0.86      0.88   \n",
       "BernoulliNB                        0.83               0.84     0.84      0.87   \n",
       "NearestCentroid                    0.83               0.81     0.81      0.86   \n",
       "DecisionTreeClassifier             0.82               0.81     0.81      0.86   \n",
       "Perceptron                         0.84               0.81     0.81      0.87   \n",
       "PassiveAggressiveClassifier        0.81               0.78     0.78      0.85   \n",
       "KNeighborsClassifier               0.80               0.76     0.76      0.84   \n",
       "ExtraTreeClassifier                0.76               0.75     0.75      0.82   \n",
       "GaussianNB                         0.92               0.58     0.58      0.91   \n",
       "QuadraticDiscriminantAnalysis      0.93               0.54     0.54      0.90   \n",
       "DummyClassifier                    0.50               0.50     0.50      0.61   \n",
       "\n",
       "                               Time Taken  \n",
       "Model                                      \n",
       "LGBMClassifier                       0.42  \n",
       "XGBClassifier                        0.77  \n",
       "RandomForestClassifier               0.91  \n",
       "LinearSVC                            0.68  \n",
       "CalibratedClassifierCV               2.11  \n",
       "SVC                                  1.93  \n",
       "LogisticRegression                   0.08  \n",
       "NuSVC                                2.46  \n",
       "AdaBoostClassifier                   0.54  \n",
       "ExtraTreesClassifier                 0.93  \n",
       "BaggingClassifier                    0.41  \n",
       "RidgeClassifierCV                    0.07  \n",
       "LinearDiscriminantAnalysis           0.08  \n",
       "RidgeClassifier                      0.05  \n",
       "SGDClassifier                        0.09  \n",
       "BernoulliNB                          0.05  \n",
       "NearestCentroid                      0.05  \n",
       "DecisionTreeClassifier               0.08  \n",
       "Perceptron                           0.06  \n",
       "PassiveAggressiveClassifier          0.05  \n",
       "KNeighborsClassifier                 3.07  \n",
       "ExtraTreeClassifier                  0.05  \n",
       "GaussianNB                           0.05  \n",
       "QuadraticDiscriminantAnalysis        0.08  \n",
       "DummyClassifier                      0.05  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "models_u"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "01032a2c",
   "metadata": {},
   "source": [
    "We see that the highest F1 score achieved is 94% when class balancing is done using SMOTE oversampling. We'll continue with SMOTE oversampling and build individual models for models with high F1 score."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f86e0300",
   "metadata": {},
   "source": [
    "# Modeling"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8ff0dad",
   "metadata": {},
   "source": [
    "Based on diversity, time taken to run and F1 score, we have selected the following models for individual modeling: LightGBM, Logistic Regression, Support Vector Machine, Stochastic Gradient Descent and Passive Aggressive Classifier. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f4d5ff6",
   "metadata": {},
   "source": [
    "### Evaluation Metric - We have selected 'Recall for class 1' as our evaluation metric because the company would like to predict maximum number of people who would subscribe to a term deposit. \n",
    "\n",
    "Recall in our case is (subscribed customers predicted correctly)/(subscribed customers predicted correctly + subscribed customers predicted incorrectly).\n",
    "\n",
    "We would like to minimize the number of subscribed customers predicted incorrectly as it would be a lost opportunity. And hence would like to maximize Recall."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "915554b1",
   "metadata": {},
   "source": [
    "### a. Light GBM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "36261ede",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.97      0.96      0.96      7424\n",
      "           1       0.54      0.60      0.57       576\n",
      "\n",
      "    accuracy                           0.93      8000\n",
      "   macro avg       0.76      0.78      0.77      8000\n",
      "weighted avg       0.94      0.93      0.94      8000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "lgbm = LGBMClassifier()                                         # Create an instance of LGBMClassifier\n",
    "lgbm.fit(X_smo, y_smo)                                          # Fit model with SMOTE oversampled training data\n",
    "print(classification_report(y_test, lgbm.predict(X_test)))      # Generate classification report"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1df44c60",
   "metadata": {},
   "source": [
    "The F1 score using LGBMClassifier is 93% which is quite good. However, Recall for class 1 is only 60%. That is, our model is able to predict customers who will not subscribe to term deposit marketing with much greater accuracy but is unable to maintain high accuracy for people who will subscribe."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "ecac90e9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.83600067, 0.9703504 , 0.9724562 , 0.97321429, 0.97456199])"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(lgbm, X_smo, y_smo, cv=5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "87f5765e",
   "metadata": {},
   "source": [
    "The 5-fold cross validation indicates that accuracy should be in the range [0.83, 0.97]. Our F1 score of 0.93 fits in this range."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e8e64baa",
   "metadata": {},
   "source": [
    "### b. Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "6d9616c6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(random_state=42)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "logreg = LogisticRegression(random_state = seed_value)           # Create an instance of LogisticRegression\n",
    "logreg.fit(X_smo, y_smo)                                         # Fit model with training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "542c8be3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.97      0.91      0.94      7424\n",
      "           1       0.34      0.59      0.43       576\n",
      "\n",
      "    accuracy                           0.89      8000\n",
      "   macro avg       0.65      0.75      0.69      8000\n",
      "weighted avg       0.92      0.89      0.90      8000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, logreg.predict(X_test)))     # Generate classification report"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "efc393a9",
   "metadata": {},
   "source": [
    "The recall for class 1 has dropped slightly to 59% compared to 60% for LightGBM."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "766f2653",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.84307615, 0.91610512, 0.91265162, 0.91391509, 0.92419137])"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(logreg, X_smo, y_smo, cv=5)                      # Generate cross validation report"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "932a4a9b",
   "metadata": {},
   "source": [
    "The F1 score of 89% fits in the cross validation range of [84%, 92%] based on 5-fold cross validation. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f1908d64",
   "metadata": {},
   "source": [
    "### c. Support Vector Machine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "0b966535",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      0.85      0.91      7424\n",
      "           1       0.28      0.76      0.41       576\n",
      "\n",
      "    accuracy                           0.84      8000\n",
      "   macro avg       0.63      0.80      0.66      8000\n",
      "weighted avg       0.93      0.84      0.87      8000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "svc = SVC()                                                     # Create an instance of SVC\n",
    "svc.fit(X_smo, y_smo)                                           # Fit model with SMOTE oversampled training data\n",
    "print(classification_report(y_test, svc.predict(X_test)))       # Generate classification report"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb85356f",
   "metadata": {},
   "source": [
    "The F1 score for SVC is only 84% compared to 93% predicted by Lazypredict for the same model. This is much lower than 93% F1 score predicted by Light GBM model. However, the recall for class 1 is 76% which is highest till now making it one of models of interest."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "69f73a80",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.79144205, 0.80121294, 0.79169474, 0.79405323, 0.79514825])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(svc, X_smo, y_smo, cv=5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ac82e21f",
   "metadata": {},
   "source": [
    "The 5-fold cross validation indicates that accuracy should be in the range [0.79, 0.80] but our F1 score falls outside this range at 84% which is surprising."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d85480f1",
   "metadata": {},
   "source": [
    "### d. Stochastic Gradient Descent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "dfb51373",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SGDClassifier(random_state=42)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sgdc = SGDClassifier(random_state = seed_value)                      # Create an instance of SGDClassifier\n",
    "sgdc.fit(X_smo, y_smo)                                               # Fit model with training data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "c9551596",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      0.76      0.86      7424\n",
      "           1       0.21      0.82      0.33       576\n",
      "\n",
      "    accuracy                           0.76      8000\n",
      "   macro avg       0.60      0.79      0.60      8000\n",
      "weighted avg       0.93      0.76      0.82      8000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, sgdc.predict(X_test)))           # Generate classification report"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "232c0b92",
   "metadata": {},
   "source": [
    "The recall for class 1 increases even further to 82%! In the meanwhile overall F1 score has dipped to 76%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "d2e0e611",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.78049191, 0.67284367, 0.70881065, 0.69541779, 0.69870283])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(sgdc, X_smo, y_smo, cv=5)                            # Generate cross validation report"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e4695166",
   "metadata": {},
   "source": [
    "The 5-fold cross validation states that the accuracy should be in the range [67%, 78%]. Our F1 score of 76% falls in this range."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "484419f7",
   "metadata": {},
   "source": [
    "### e. Passive Aggressive Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "5c21e869",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.97      0.90      0.94      7424\n",
      "           1       0.35      0.65      0.45       576\n",
      "\n",
      "    accuracy                           0.89      8000\n",
      "   macro avg       0.66      0.78      0.69      8000\n",
      "weighted avg       0.93      0.89      0.90      8000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "pac = PassiveAggressiveClassifier(random_state = seed_value)           # Create an instance of PassiveAggressiveClassifier\n",
    "pac.fit(X_smo, y_smo)                                                  # Fit model with training data\n",
    "print(classification_report(y_test, pac.predict(X_test)))              # Generate classification report"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a1a93be",
   "metadata": {},
   "source": [
    "The recall for class 1 is third highest at 65%. The F1 score is also decent at 89%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "2c6679af",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.76314016, 0.91492588, 0.84661388, 0.73803908, 0.66399933])"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cross_val_score(pac, X_smo, y_smo, cv=5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "32b47d86",
   "metadata": {},
   "source": [
    "The 5-fold cross validation suggests that accuracy should be in the range [0.84, 0.98]. Our F1 score of 93% falls in this range."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5721afd3",
   "metadata": {},
   "source": [
    "Based on the modeling so far, Stochastic Gradient Descent, Support Vector Machine and Passive Aggressive Classifer give the highest recall values for class 1. Lets build a ensemble model based on the above models to make our final model robust."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1439b9d3",
   "metadata": {},
   "source": [
    "# Ensemble Modeling"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c14f44d3",
   "metadata": {},
   "source": [
    "We'll experiment with two types of ensemble modeling to see if that increases our accuracy. The two types of ensemble models are: Stacking Classifier and Voting Classifier."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d73e5c43",
   "metadata": {},
   "source": [
    "### a. Stacking Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "62501ea4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a list of initial estimators for Stacking Classifier\n",
    "estimators = [('svc_em', SVC(random_state = seed_value)), ('pac_em', PassiveAggressiveClassifier(random_state = seed_value))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "509656ab",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      0.87      0.92      7424\n",
      "           1       0.31      0.74      0.44       576\n",
      "\n",
      "    accuracy                           0.86      8000\n",
      "   macro avg       0.64      0.81      0.68      8000\n",
      "weighted avg       0.93      0.86      0.89      8000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "stc = StackingClassifier(estimators = estimators, final_estimator = SGDClassifier(random_state = seed_value))  # Create an instance of Stacking Classifier\n",
    "stc.fit(X_smo, y_smo)                                               # Fit model with training data\n",
    "print(classification_report(y_test, stc.predict(X_test)))           # Generate classification report"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e01f3d6",
   "metadata": {},
   "source": [
    "The highest recall for class 1 after experimenting with final estimator is 74%."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e70b99e",
   "metadata": {},
   "source": [
    "### b. Voting Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "726e5229",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build a voting classifier\n",
    "vot_clf = VotingClassifier(estimators = [('sgdc', sgdc), ('svc', svc), ('pac', pac)], voting = 'hard')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "4406c919",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.98      0.85      0.91      7424\n",
      "           1       0.28      0.76      0.41       576\n",
      "\n",
      "    accuracy                           0.84      8000\n",
      "   macro avg       0.63      0.80      0.66      8000\n",
      "weighted avg       0.93      0.84      0.87      8000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "vot_clf.fit(X_smo, y_smo)\n",
    "print(classification_report(y_test, vot_clf.predict(X_test)))                # Generate classification report"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67e691ca",
   "metadata": {},
   "source": [
    "The recall for class 1 using VotingClassifier is 76% which is quite good. However, it is lower than 82% recall achieved using SGDClassifier."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e63b2d49",
   "metadata": {},
   "source": [
    "Based on the modeling so far, we see that SGD model gives the highest recall for class 1 at 82%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "c1ce12ea",
   "metadata": {},
   "outputs": [],
   "source": [
    "# hierarchical clustering - diff colors, kmeans, shap feature importance"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f190a75b",
   "metadata": {},
   "source": [
    "# Feature Importance"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de3af69e",
   "metadata": {},
   "source": [
    "We'll now identify important features to understand what makes the customer subscribe to the term deposit marketing."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e389ae0a",
   "metadata": {},
   "source": [
    "One way to understand importance of features is by calculating magnitude of coefficient times the standard deviation for the parameters. We'll consider SGD model as it gave us the highest recall for class 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "1ac72bd4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-7.32255621e+01,  1.94061565e+00, -3.64016260e+01,\n",
       "         2.42375857e+01, -6.86027743e+02, -2.93513090e+02,\n",
       "        -4.57050179e+01, -3.07205491e+01, -1.63708024e+02,\n",
       "        -2.13995609e+01, -4.38875760e+01, -1.24586519e+02,\n",
       "        -1.30664769e+01, -1.72616757e+02, -3.63991122e+01,\n",
       "        -7.58650005e+00, -3.89060760e+02, -1.97399826e+02,\n",
       "        -4.16111232e+02, -1.25456277e+02, -4.70398196e+01,\n",
       "        -1.71211570e+01, -6.10419145e+02, -2.07733856e+02,\n",
       "        -4.79146049e+01, -4.28539217e+02, -1.27386838e+02,\n",
       "        -4.02200135e-01, -3.43001303e+01, -3.99334459e+01,\n",
       "        -1.86620863e+02, -1.49497790e+02,  2.00748143e+01,\n",
       "        -3.93784097e+02, -1.33246391e+02,  5.93496575e+00]])"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Estimated coefficients for SGDClassifier\n",
    "sgdc.coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "0001ba18",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[-7.06015606e+02  5.63503052e+03 -3.01333527e+02  6.28633913e+03\n",
      "  -2.22205141e+03 -1.24371152e+02 -8.41409820e+00 -4.99494961e+00\n",
      "  -6.59872863e+01 -3.98252639e+00 -8.10440597e+00 -3.69992467e+01\n",
      "  -1.48569864e+00 -6.50369706e+01 -5.96304117e+00 -5.79783562e-01\n",
      "  -1.89795019e+02 -8.78636693e+01 -2.07799017e+02 -5.63389455e+01\n",
      "  -9.02503095e+00 -2.41013046e+00 -2.98946131e+02 -7.86196319e+01\n",
      "  -1.12019898e+01 -1.99758229e+02 -4.28931287e+01 -7.24958771e-03\n",
      "  -7.97839316e+00 -6.74575551e+00 -6.83296814e+01 -4.82910903e+01\n",
      "   1.60703841e+00 -1.86311533e+02 -3.81231402e+01  2.65154184e-01]]\n"
     ]
    }
   ],
   "source": [
    "# Feature importance - Normalizing relationship between coefficients\n",
    "print(np.std(X, 0).values*sgdc.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "b0e3166a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['age', 'balance', 'day', 'duration', 'campaign', 'job_blue-collar',\n",
       "       'job_entrepreneur', 'job_housemaid', 'job_management', 'job_retired',\n",
       "       'job_self-employed', 'job_services', 'job_student', 'job_technician',\n",
       "       'job_unemployed', 'job_unknown', 'marital_married', 'marital_single',\n",
       "       'education_secondary', 'education_tertiary', 'education_unknown',\n",
       "       'default_yes', 'housing_yes', 'loan_yes', 'contact_telephone',\n",
       "       'contact_unknown', 'month_aug', 'month_dec', 'month_feb', 'month_jan',\n",
       "       'month_jul', 'month_jun', 'month_mar', 'month_may', 'month_nov',\n",
       "       'month_oct'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e02479a9",
   "metadata": {},
   "source": [
    "From the above feature importance, we see that duration and balance are most positively influential parameters whereas campaign is most negatively influential parameter. That is people who are more likely to convert have higher balance and high talk time in the last call."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f24d92b",
   "metadata": {},
   "source": [
    "# Clustering"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f1050f5",
   "metadata": {},
   "source": [
    "We'll now use Kmeans clustering technique to identify customers who are more likely to subscribe to the term deposit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "df089299",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Since Kmeans is a distance-based algorithm, we need to standarize the data\n",
    "scaler = StandardScaler()                              # Create an instance of StandardScaler\n",
    "tdm_scale = scaler.fit_transform(X)                    # Standardize data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "5b1aeca6",
   "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>age</th>\n",
       "      <th>balance</th>\n",
       "      <th>day</th>\n",
       "      <th>duration</th>\n",
       "      <th>campaign</th>\n",
       "      <th>job_blue-collar</th>\n",
       "      <th>job_entrepreneur</th>\n",
       "      <th>job_housemaid</th>\n",
       "      <th>job_management</th>\n",
       "      <th>job_retired</th>\n",
       "      <th>...</th>\n",
       "      <th>month_aug</th>\n",
       "      <th>month_dec</th>\n",
       "      <th>month_feb</th>\n",
       "      <th>month_jan</th>\n",
       "      <th>month_jul</th>\n",
       "      <th>month_jun</th>\n",
       "      <th>month_mar</th>\n",
       "      <th>month_may</th>\n",
       "      <th>month_nov</th>\n",
       "      <th>month_oct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.81</td>\n",
       "      <td>0.30</td>\n",
       "      <td>-1.33</td>\n",
       "      <td>0.02</td>\n",
       "      <td>-0.58</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>1.97</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.39</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.37</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>1.40</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.36</td>\n",
       "      <td>-0.43</td>\n",
       "      <td>-1.33</td>\n",
       "      <td>-0.40</td>\n",
       "      <td>-0.58</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.51</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.39</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.37</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>1.40</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.78</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-1.33</td>\n",
       "      <td>-0.69</td>\n",
       "      <td>-0.58</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>5.24</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.51</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.39</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.37</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>1.40</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.67</td>\n",
       "      <td>0.08</td>\n",
       "      <td>-1.33</td>\n",
       "      <td>-0.63</td>\n",
       "      <td>-0.58</td>\n",
       "      <td>1.81</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.51</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.39</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.37</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>1.40</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.78</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-1.33</td>\n",
       "      <td>-0.22</td>\n",
       "      <td>-0.58</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.51</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.39</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.37</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>1.40</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.04</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 36 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    age  balance   day  duration  campaign  job_blue-collar  job_entrepreneur  \\\n",
       "0  1.81     0.30 -1.33      0.02     -0.58            -0.55             -0.19   \n",
       "1  0.36    -0.43 -1.33     -0.40     -0.58            -0.55             -0.19   \n",
       "2 -0.78    -0.44 -1.33     -0.69     -0.58            -0.55              5.24   \n",
       "3  0.67     0.08 -1.33     -0.63     -0.58             1.81             -0.19   \n",
       "4 -0.78    -0.44 -1.33     -0.22     -0.58            -0.55             -0.19   \n",
       "\n",
       "   job_housemaid  job_management  job_retired  ...  month_aug  month_dec  \\\n",
       "0          -0.17            1.97        -0.19  ...      -0.39      -0.02   \n",
       "1          -0.17           -0.51        -0.19  ...      -0.39      -0.02   \n",
       "2          -0.17           -0.51        -0.19  ...      -0.39      -0.02   \n",
       "3          -0.17           -0.51        -0.19  ...      -0.39      -0.02   \n",
       "4          -0.17           -0.51        -0.19  ...      -0.39      -0.02   \n",
       "\n",
       "   month_feb  month_jan  month_jul  month_jun  month_mar  month_may  \\\n",
       "0      -0.25      -0.17      -0.44      -0.37      -0.08       1.40   \n",
       "1      -0.25      -0.17      -0.44      -0.37      -0.08       1.40   \n",
       "2      -0.25      -0.17      -0.44      -0.37      -0.08       1.40   \n",
       "3      -0.25      -0.17      -0.44      -0.37      -0.08       1.40   \n",
       "4      -0.25      -0.17      -0.44      -0.37      -0.08       1.40   \n",
       "\n",
       "   month_nov  month_oct  \n",
       "0      -0.31      -0.04  \n",
       "1      -0.31      -0.04  \n",
       "2      -0.31      -0.04  \n",
       "3      -0.31      -0.04  \n",
       "4      -0.31      -0.04  \n",
       "\n",
       "[5 rows x 36 columns]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tdm_scaled = pd.DataFrame(tdm_scale, columns = X.columns)        # Create a dataframe with standarized data\n",
    "tdm_scaled.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "b19a7fa9",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>balance</th>\n",
       "      <th>day</th>\n",
       "      <th>duration</th>\n",
       "      <th>campaign</th>\n",
       "      <th>job_blue-collar</th>\n",
       "      <th>job_entrepreneur</th>\n",
       "      <th>job_housemaid</th>\n",
       "      <th>job_management</th>\n",
       "      <th>job_retired</th>\n",
       "      <th>...</th>\n",
       "      <th>month_aug</th>\n",
       "      <th>month_dec</th>\n",
       "      <th>month_feb</th>\n",
       "      <th>month_jan</th>\n",
       "      <th>month_jul</th>\n",
       "      <th>month_jun</th>\n",
       "      <th>month_mar</th>\n",
       "      <th>month_may</th>\n",
       "      <th>month_nov</th>\n",
       "      <th>month_oct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>...</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "      <td>40000.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>-0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>...</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-2.23</td>\n",
       "      <td>-3.20</td>\n",
       "      <td>-1.81</td>\n",
       "      <td>-0.98</td>\n",
       "      <td>-0.58</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.51</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.39</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.37</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>-0.72</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.78</td>\n",
       "      <td>-0.42</td>\n",
       "      <td>-0.97</td>\n",
       "      <td>-0.60</td>\n",
       "      <td>-0.58</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.51</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.39</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.37</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>-0.72</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>-0.16</td>\n",
       "      <td>-0.30</td>\n",
       "      <td>0.12</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.27</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.51</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.39</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.37</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>-0.72</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.77</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.60</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.04</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.51</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.39</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.25</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.37</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>1.40</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>-0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>5.65</td>\n",
       "      <td>34.73</td>\n",
       "      <td>1.81</td>\n",
       "      <td>17.98</td>\n",
       "      <td>18.56</td>\n",
       "      <td>1.81</td>\n",
       "      <td>5.24</td>\n",
       "      <td>5.98</td>\n",
       "      <td>1.97</td>\n",
       "      <td>5.18</td>\n",
       "      <td>...</td>\n",
       "      <td>2.58</td>\n",
       "      <td>55.46</td>\n",
       "      <td>4.05</td>\n",
       "      <td>5.75</td>\n",
       "      <td>2.30</td>\n",
       "      <td>2.73</td>\n",
       "      <td>12.41</td>\n",
       "      <td>1.40</td>\n",
       "      <td>3.18</td>\n",
       "      <td>22.34</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 36 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           age  balance      day  duration  campaign  job_blue-collar  \\\n",
       "count 40000.00 40000.00 40000.00  40000.00  40000.00         40000.00   \n",
       "mean     -0.00    -0.00     0.00      0.00     -0.00             0.00   \n",
       "std       1.00     1.00     1.00      1.00      1.00             1.00   \n",
       "min      -2.23    -3.20    -1.81     -0.98     -0.58            -0.55   \n",
       "25%      -0.78    -0.42    -0.97     -0.60     -0.58            -0.55   \n",
       "50%      -0.16    -0.30     0.12     -0.31     -0.27            -0.55   \n",
       "75%       0.77     0.02     0.60      0.22      0.04            -0.55   \n",
       "max       5.65    34.73     1.81     17.98     18.56             1.81   \n",
       "\n",
       "       job_entrepreneur  job_housemaid  job_management  job_retired  ...  \\\n",
       "count          40000.00       40000.00        40000.00     40000.00  ...   \n",
       "mean              -0.00           0.00           -0.00         0.00  ...   \n",
       "std                1.00           1.00            1.00         1.00  ...   \n",
       "min               -0.19          -0.17           -0.51        -0.19  ...   \n",
       "25%               -0.19          -0.17           -0.51        -0.19  ...   \n",
       "50%               -0.19          -0.17           -0.51        -0.19  ...   \n",
       "75%               -0.19          -0.17           -0.51        -0.19  ...   \n",
       "max                5.24           5.98            1.97         5.18  ...   \n",
       "\n",
       "       month_aug  month_dec  month_feb  month_jan  month_jul  month_jun  \\\n",
       "count   40000.00   40000.00   40000.00   40000.00   40000.00   40000.00   \n",
       "mean        0.00       0.00      -0.00       0.00       0.00       0.00   \n",
       "std         1.00       1.00       1.00       1.00       1.00       1.00   \n",
       "min        -0.39      -0.02      -0.25      -0.17      -0.44      -0.37   \n",
       "25%        -0.39      -0.02      -0.25      -0.17      -0.44      -0.37   \n",
       "50%        -0.39      -0.02      -0.25      -0.17      -0.44      -0.37   \n",
       "75%        -0.39      -0.02      -0.25      -0.17      -0.44      -0.37   \n",
       "max         2.58      55.46       4.05       5.75       2.30       2.73   \n",
       "\n",
       "       month_mar  month_may  month_nov  month_oct  \n",
       "count   40000.00   40000.00   40000.00   40000.00  \n",
       "mean        0.00       0.00      -0.00       0.00  \n",
       "std         1.00       1.00       1.00       1.00  \n",
       "min        -0.08      -0.72      -0.31      -0.04  \n",
       "25%        -0.08      -0.72      -0.31      -0.04  \n",
       "50%        -0.08      -0.72      -0.31      -0.04  \n",
       "75%        -0.08       1.40      -0.31      -0.04  \n",
       "max        12.41       1.40       3.18      22.34  \n",
       "\n",
       "[8 rows x 36 columns]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tdm_scaled.describe()                  # Get a summary of the standarized dataframe"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0d98737d",
   "metadata": {},
   "source": [
    "The maginitude of all the variables is similar after scaling."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "298cf5f5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Inertia')"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# fitting multiple k-means algorithms and storing the values in an empty list\n",
    "SSE = []\n",
    "for cluster in range(1,50):\n",
    "    kmeans = KMeans(n_jobs = -1, n_clusters = cluster, init='k-means++')\n",
    "    kmeans.fit(tdm_scaled)\n",
    "    SSE.append(kmeans.inertia_)\n",
    "\n",
    "# converting the results into a dataframe and plotting them\n",
    "frame = pd.DataFrame({'Cluster':range(1,50), 'SSE':SSE})\n",
    "plt.figure(figsize=(12,6))\n",
    "plt.plot(frame['Cluster'], frame['SSE'], marker='o')\n",
    "plt.xlabel('Number of clusters')\n",
    "plt.ylabel('Inertia')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "6a41495b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KMeans(n_clusters=23)"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Defining the K-means cluster with centroid initialization using Kmeans++. \n",
    "# Kmeans++ helps in finding an optimized imitial centroid.\n",
    "kmeans2 = KMeans(n_clusters = 23, init ='k-means++')\n",
    "kmeans2.fit(tdm_scaled)                                          # Fit scaled data in Kmeans algorithm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "0a74d2b6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "650184.233602959"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "kmeans2.inertia_"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d5a6ce61",
   "metadata": {},
   "source": [
    "The elbow occures at k = 23 and the intertia at k = 23 is 650184. This is the total number of clusters in the dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bf8415f6",
   "metadata": {},
   "source": [
    "# Principal Component Analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "37391d68",
   "metadata": {},
   "source": [
    "We currently have 30+ variables in the dataset. We now want to see whether we can reduce the dimensions so that it is easier to interpret clusters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "10e652a6",
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>13</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "      <th>17</th>\n",
       "      <th>18</th>\n",
       "      <th>19</th>\n",
       "      <th>20</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.60</td>\n",
       "      <td>-0.03</td>\n",
       "      <td>3.50</td>\n",
       "      <td>-0.94</td>\n",
       "      <td>-0.54</td>\n",
       "      <td>-0.88</td>\n",
       "      <td>-0.13</td>\n",
       "      <td>-0.12</td>\n",
       "      <td>1.32</td>\n",
       "      <td>0.11</td>\n",
       "      <td>...</td>\n",
       "      <td>0.41</td>\n",
       "      <td>0.03</td>\n",
       "      <td>-0.10</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.14</td>\n",
       "      <td>-0.49</td>\n",
       "      <td>-0.18</td>\n",
       "      <td>-0.33</td>\n",
       "      <td>0.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-2.17</td>\n",
       "      <td>1.90</td>\n",
       "      <td>-0.06</td>\n",
       "      <td>1.17</td>\n",
       "      <td>0.39</td>\n",
       "      <td>-1.31</td>\n",
       "      <td>0.30</td>\n",
       "      <td>-0.15</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.37</td>\n",
       "      <td>...</td>\n",
       "      <td>0.31</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.11</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>0.55</td>\n",
       "      <td>0.34</td>\n",
       "      <td>-0.77</td>\n",
       "      <td>-0.60</td>\n",
       "      <td>0.32</td>\n",
       "      <td>0.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-2.18</td>\n",
       "      <td>-0.34</td>\n",
       "      <td>0.96</td>\n",
       "      <td>-0.90</td>\n",
       "      <td>-0.86</td>\n",
       "      <td>-0.17</td>\n",
       "      <td>-1.38</td>\n",
       "      <td>-1.55</td>\n",
       "      <td>-0.36</td>\n",
       "      <td>0.24</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.34</td>\n",
       "      <td>-0.42</td>\n",
       "      <td>-2.25</td>\n",
       "      <td>0.08</td>\n",
       "      <td>-0.58</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>2.28</td>\n",
       "      <td>0.45</td>\n",
       "      <td>1.39</td>\n",
       "      <td>-1.37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-2.03</td>\n",
       "      <td>-1.01</td>\n",
       "      <td>2.66</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-0.16</td>\n",
       "      <td>1.21</td>\n",
       "      <td>1.15</td>\n",
       "      <td>2.82</td>\n",
       "      <td>-0.46</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.48</td>\n",
       "      <td>0.08</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>-0.51</td>\n",
       "      <td>1.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.73</td>\n",
       "      <td>1.12</td>\n",
       "      <td>1.93</td>\n",
       "      <td>3.59</td>\n",
       "      <td>1.40</td>\n",
       "      <td>4.37</td>\n",
       "      <td>2.79</td>\n",
       "      <td>4.27</td>\n",
       "      <td>-1.00</td>\n",
       "      <td>1.19</td>\n",
       "      <td>...</td>\n",
       "      <td>2.01</td>\n",
       "      <td>0.82</td>\n",
       "      <td>3.20</td>\n",
       "      <td>0.53</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.46</td>\n",
       "      <td>-2.43</td>\n",
       "      <td>0.18</td>\n",
       "      <td>2.96</td>\n",
       "      <td>-4.42</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     0     1     2     3     4     5     6     7     8    9   ...    13    14  \\\n",
       "0  0.60 -0.03  3.50 -0.94 -0.54 -0.88 -0.13 -0.12  1.32 0.11  ...  0.41  0.03   \n",
       "1 -2.17  1.90 -0.06  1.17  0.39 -1.31  0.30 -0.15  0.78 0.37  ...  0.31 -0.02   \n",
       "2 -2.18 -0.34  0.96 -0.90 -0.86 -0.17 -1.38 -1.55 -0.36 0.24  ... -1.34 -0.42   \n",
       "3 -2.03 -1.01  2.66 -0.19 -0.16  1.21  1.15  2.82 -0.46 0.03  ...  0.07  0.12   \n",
       "4 -0.73  1.12  1.93  3.59  1.40  4.37  2.79  4.27 -1.00 1.19  ...  2.01  0.82   \n",
       "\n",
       "     15    16    17    18    19    20    21    22  \n",
       "0 -0.10 -0.08  0.13  0.14 -0.49 -0.18 -0.33  0.08  \n",
       "1 -0.11 -0.19  0.55  0.34 -0.77 -0.60  0.32  0.15  \n",
       "2 -2.25  0.08 -0.58 -0.21  2.28  0.45  1.39 -1.37  \n",
       "3  0.10  0.17  0.48  0.08 -0.44 -0.44 -0.51  1.06  \n",
       "4  3.20  0.53  0.22  0.46 -2.43  0.18  2.96 -4.42  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca=PCA(n_components=23)                                 # Initialize PCA with number of components\n",
    "principalComponents=pca.fit_transform(tdm_scaled)        # Fit with training data\n",
    "principalDf=pd.DataFrame(data=principalComponents)       # Create a dataframe\n",
    "principalDf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "7bf3ff65",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.07445064, 0.06389086, 0.05504755, 0.0457614 , 0.04101336,\n",
       "       0.03862569, 0.03708491, 0.03593419, 0.03282187, 0.03173509,\n",
       "       0.03130772, 0.03118258, 0.02961049, 0.02915454, 0.02898882,\n",
       "       0.02876742, 0.02819299, 0.02794006, 0.02768792, 0.02728425,\n",
       "       0.02692807, 0.02544978, 0.02435645])"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca.explained_variance_ratio_                 # Get percentage explained value for each principal component"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "68de9592",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8232166375542591"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca.explained_variance_ratio_.sum()            # Total percentage explained by 23 principal components"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "777c2eed",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Percentage Variance Explained')"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Generate variance explained for multiple values of principal components and storing the values in an empty list\n",
    "perc_var_exp = []\n",
    "for n_component in range(1,36):\n",
    "    pca2=PCA(n_components = n_component) \n",
    "    principalComponents=pca2.fit_transform(tdm_scaled) \n",
    "    principalDf=pd.DataFrame(data=principalComponents) \n",
    "    perc_var_exp.append(pca2.explained_variance_ratio_.sum())\n",
    "    \n",
    "# converting the results into a dataframe and plotting them\n",
    "frame2 = pd.DataFrame({'Principal_Components':range(1,36), 'Percentage_Variance_Explained': perc_var_exp})\n",
    "plt.figure(figsize=(12,6))\n",
    "plt.plot(frame2['Principal_Components'], frame2['Percentage_Variance_Explained'], marker='o')\n",
    "plt.xlabel('Number of Principal Components')\n",
    "plt.ylabel('Percentage Variance Explained')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60cb3c47",
   "metadata": {},
   "source": [
    "23 principal components are required to explain 82.32 % of variance in the dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8580bae",
   "metadata": {},
   "source": [
    "# Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9ebf2f7e",
   "metadata": {},
   "source": [
    "The objective of the project was to predict whether a customer will subscribe to a term deposit. Since we want to make sure that we don't miss out on revenue from incorrectly classified people who will actually subscribe, hence we chose the evaluation metric as Recall for class 1. Using Stochaistic Gradient Descent, we got the maximum recall of 82% for class 1. \n",
    "\n",
    "Using feature importance, we note that duration and balance are the most important features which make the customer subscribe to a term deposit. That is, longer the last call with a potential subscriber and higher the balance, greater the chances for that person subscribing to a term deposit.\n",
    "\n",
    "Using K-means clustering, we see that there are 23 different groups of people in the dataset. Hence, it is difficult to analyze which groups of people are likely to subscribe to a term deposit.\n",
    "\n",
    "23 principal components are required to explain 82.32 % of variance in the dataset. That is, it is not possible to reduce the dimension significantly for this dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a86b2456",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "vtdm",
   "language": "python",
   "name": "vtdm"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}