loan_prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Binary Classification Machine Learning. Case Study Loan Prediction\n",
    "\n",
    "## 1 Problem Definition\n",
    "\n",
    "The aim of this exercise is to use **Machine Learning** techniques to predict loan eligibility based on customer details. These details are numerical and categorical data that include information about gender, marital status, education, dependents, income, loan amount, credit history, and others. The label of each record is the letter `Y` if the loan is approved and `N` is the loan is rejected.\n",
    "\n",
    "The training dataset contains the following information:\n",
    "\n",
    "<table><tbody><tr><td width=\"155\"><p><strong>Variable</strong></p></td><td width=\"465\"><p>Description</p></td></tr><tr><td width=\"155\"><p><strong>Loan_ID</strong></p></td><td width=\"465\"><p>Unique Loan ID</p></td></tr><tr><td width=\"155\"><p><strong>Gender</strong></p></td><td width=\"465\"><p>Male/ Female</p></td></tr><tr><td width=\"155\"><p><strong>Married</strong></p></td><td width=\"465\"><p>Applicant married (Y/N)</p></td></tr><tr><td width=\"155\"><p><strong>Dependents</strong></p></td><td width=\"465\"><p>Number of dependents</p></td></tr><tr><td width=\"155\"><p><strong>Education</strong></p></td><td width=\"465\"><p>Applicant Education (Graduate/ Under Graduate)</p></td></tr><tr><td width=\"155\"><p><strong>Self_Employed</strong></p></td><td width=\"465\"><p>Self employed (Y/N)</p></td></tr><tr><td width=\"155\"><p><strong>ApplicantIncome</strong></p></td><td width=\"465\"><p>Applicant income</p></td></tr><tr><td width=\"155\"><p><strong>CoapplicantIncome</strong></p></td><td width=\"465\"><p>Coapplicant income</p></td></tr><tr><td width=\"155\"><p><strong>LoanAmount</strong></p></td><td width=\"465\"><p>Loan amount in thousands</p></td></tr><tr><td width=\"155\"><p><strong>Loan_Amount_Term</strong></p></td><td width=\"465\"><p>Term of loan in months</p></td></tr><tr><td width=\"155\"><p><strong>Credit_History</strong></p></td><td width=\"465\"><p>credit history meets guidelines</p></td></tr><tr><td width=\"155\"><p><strong>Property_Area</strong></p></td><td width=\"465\"><p>Urban/ Semi Urban/ Rural</p></td></tr><tr><td width=\"155\"><p><strong>Loan_Status</strong></p></td><td width=\"465\"><p>Loan approved (Y/N)</p></td></tr></tbody></table><p> \n",
    "    \n",
    "The test dataset contains the same columns as the training dataset except the last one (Loan approved).\n",
    "\n",
    "To solve this problem we will implement a Machine Learning solution based on data transformation and algorithm tuning to improve model performance. This solution is inspired by the recipes contained in the book [**Machine Learning Mastery With Python**](https://machinelearningmastery.com/machine-learning-with-python/) by Jason Brownlee [1]."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2 Load the Dataset\n",
    "\n",
    "First, we will load the required libraries to accomplish our goals."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load libraries\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "from pandas import set_option\n",
    "from pandas.plotting import scatter_matrix\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "from sklearn.pipeline import Pipeline\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble import ExtraTreesClassifier\n",
    "\n",
    "from sklearn.impute import SimpleImputer\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.1 Loading training dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>Loan_ID</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Married</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>Education</th>\n",
       "      <th>Self_Employed</th>\n",
       "      <th>ApplicantIncome</th>\n",
       "      <th>CoapplicantIncome</th>\n",
       "      <th>LoanAmount</th>\n",
       "      <th>Loan_Amount_Term</th>\n",
       "      <th>Credit_History</th>\n",
       "      <th>Property_Area</th>\n",
       "      <th>Loan_Status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LP001002</td>\n",
       "      <td>Male</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>No</td>\n",
       "      <td>5849</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Urban</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LP001003</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>1</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>No</td>\n",
       "      <td>4583</td>\n",
       "      <td>1508.0</td>\n",
       "      <td>128.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Rural</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LP001005</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>Yes</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Urban</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>LP001006</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Not Graduate</td>\n",
       "      <td>No</td>\n",
       "      <td>2583</td>\n",
       "      <td>2358.0</td>\n",
       "      <td>120.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Urban</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LP001008</td>\n",
       "      <td>Male</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>No</td>\n",
       "      <td>6000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>141.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Urban</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Loan_ID Gender Married Dependents     Education Self_Employed  \\\n",
       "0  LP001002   Male      No          0      Graduate            No   \n",
       "1  LP001003   Male     Yes          1      Graduate            No   \n",
       "2  LP001005   Male     Yes          0      Graduate           Yes   \n",
       "3  LP001006   Male     Yes          0  Not Graduate            No   \n",
       "4  LP001008   Male      No          0      Graduate            No   \n",
       "\n",
       "   ApplicantIncome  CoapplicantIncome  LoanAmount  Loan_Amount_Term  \\\n",
       "0             5849                0.0         NaN             360.0   \n",
       "1             4583             1508.0       128.0             360.0   \n",
       "2             3000                0.0        66.0             360.0   \n",
       "3             2583             2358.0       120.0             360.0   \n",
       "4             6000                0.0       141.0             360.0   \n",
       "\n",
       "   Credit_History Property_Area Loan_Status  \n",
       "0             1.0         Urban           Y  \n",
       "1             1.0         Rural           N  \n",
       "2             1.0         Urban           Y  \n",
       "3             1.0         Urban           Y  \n",
       "4             1.0         Urban           Y  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# load dataset\n",
    "df = pd.read_csv('train.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "According to the output, some columns have categorical data and other columns have numerical data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Imputation of missing values for categorical and numerical values\n",
    "\n",
    "As we do not want to reduce the amount of data for our Machine Learning solution, we need to fill null values with the most frequent one for each category in our data frame. But first, we will sum missing values for each column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Loan_ID               0\n",
       "Gender               13\n",
       "Married               3\n",
       "Dependents           15\n",
       "Education             0\n",
       "Self_Employed        32\n",
       "ApplicantIncome       0\n",
       "CoapplicantIncome     0\n",
       "LoanAmount           22\n",
       "Loan_Amount_Term     14\n",
       "Credit_History       50\n",
       "Property_Area         0\n",
       "Loan_Status           0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sum missing values \n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have seven columns with missing values. Next, we will impute both categorical and numerical values and change the data type of some columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "imp = SimpleImputer(missing_values=np.nan, strategy='most_frequent')\n",
    "df = pd.DataFrame(imp.fit_transform(df),columns=df.columns,index=df.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.astype({'ApplicantIncome': np.int64, 'CoapplicantIncome':float,\n",
    "                'LoanAmount':float,'Loan_Amount_Term':float, 'Credit_History':float})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Transforming Categorical Data\n",
    "\n",
    "In this section, we will transform categorical data because Machine Learning algorithm cannot deal with this type of data. \n",
    "\n",
    "First, we will get unique values from categorical columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Male' 'Female'] ['No' 'Yes'] ['0' '1' '2' '3+'] ['Graduate' 'Not Graduate'] ['No' 'Yes'] ['Urban' 'Rural' 'Semiurban'] ['Y' 'N']\n"
     ]
    }
   ],
   "source": [
    "print(df.Gender.unique(),\n",
    "df.Married.unique(),\n",
    "df.Dependents.unique(),\n",
    "df.Education.unique(),\n",
    "df.Self_Employed.unique(),\n",
    "df.Property_Area.unique(),\n",
    "df.Loan_Status.unique())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we will create and apply a function to generate the encoding scheme for categorical data and write back labels to the DataFrame. Based on [Dipanjan (DJ) Sarkar post]((https://towardsdatascience.com/understanding-feature-engineering-part-2-categorical-data-f54324193e63))."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import LabelEncoder\n",
    "def encoding_schema(df, cat, label):\n",
    "    \n",
    "    '''Function to generate the encoding scheme for categorical data and \n",
    "        write back labels to the DataFrame\n",
    "    '''\n",
    "    gle = LabelEncoder()\n",
    "    cat_labels = gle.fit_transform(df[cat])\n",
    "    cat_mappings = {index: label for index, label in enumerate(gle.classes_)}\n",
    "    \n",
    "    df[label] = cat_labels.astype('int64')\n",
    "    \n",
    "    return ('Labels for {} are {}'.format(cat, cat_mappings))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Gender are {0: 'Female', 1: 'Male'}\""
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Gender \n",
    "encoding_schema(df, 'Gender', 'GenderLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Married are {0: 'No', 1: 'Yes'}\""
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Married \n",
    "encoding_schema(df, 'Married', 'MarriedLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Dependents are {0: '0', 1: '1', 2: '2', 3: '3+'}\""
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Dependents\n",
    "encoding_schema(df, 'Dependents', 'DependentsLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Education are {0: 'Graduate', 1: 'Not Graduate'}\""
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Education\n",
    "encoding_schema(df, 'Education', 'EducationLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Self_Employed are {0: 'No', 1: 'Yes'}\""
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Self_Employed\n",
    "encoding_schema(df, 'Self_Employed', 'Self_EmployedLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Property_Area are {0: 'Rural', 1: 'Semiurban', 2: 'Urban'}\""
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Property_Area\n",
    "encoding_schema(df, 'Property_Area', 'Property_AreaLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Loan_Status are {0: 'N', 1: 'Y'}\""
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Loan_Status\n",
    "encoding_schema(df, 'Loan_Status', 'Loan_StatusLabel')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we will get rid of unuseful columns and rename back useful columns  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns_to_keep  = ['Loan_ID', 'GenderLabel', 'MarriedLabel', 'DependentsLabel', 'EducationLabel',\n",
    "       'Self_EmployedLabel', 'ApplicantIncome', 'CoapplicantIncome', 'LoanAmount',\n",
    "       'Loan_Amount_Term', 'Credit_History', 'Property_AreaLabel', 'Loan_StatusLabel']\n",
    "\n",
    "df = df[columns_to_keep]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "old_colnames  = ['GenderLabel', 'MarriedLabel', 'DependentsLabel', 'EducationLabel',\n",
    "       'Self_EmployedLabel', 'Property_AreaLabel', 'Loan_StatusLabel']\n",
    "\n",
    "new_colnames  = ['Gender', 'Married', 'Dependents', 'Education',\n",
    "       'Self_Employed', 'Property_Area', 'Loan_Status']\n",
    "\n",
    "col_rename_dict = {i:j for i,j in zip(old_colnames,new_colnames)}\n",
    "df = df.rename(columns=col_rename_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5 Analyzing Data\n",
    "\n",
    "After transforming our dataset, we can examine closely our data. \n",
    "\n",
    "### 5.1 Descriptive Statistics\n",
    "\n",
    "Let's examine the number of rows and columns of our dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(614, 13)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# shape\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have 614 samples to produce our solution and we can observe that our dataset contains 13 attributes including the target attribute Loan_Status.\n",
    "\n",
    "Now, let's look at the data type of each column.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Loan_ID               object\n",
       "Gender                 int64\n",
       "Married                int64\n",
       "Dependents             int64\n",
       "Education              int64\n",
       "Self_Employed          int64\n",
       "ApplicantIncome        int64\n",
       "CoapplicantIncome    float64\n",
       "LoanAmount           float64\n",
       "Loan_Amount_Term     float64\n",
       "Credit_History       float64\n",
       "Property_Area          int64\n",
       "Loan_Status            int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nearly all columns are numeric, some of them are real values (float) and others are integers (int).\n",
    "\n",
    "Next, let's take a pick at the first 5 rows."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>Loan_ID</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Married</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>Education</th>\n",
       "      <th>Self_Employed</th>\n",
       "      <th>ApplicantIncome</th>\n",
       "      <th>CoapplicantIncome</th>\n",
       "      <th>LoanAmount</th>\n",
       "      <th>Loan_Amount_Term</th>\n",
       "      <th>Credit_History</th>\n",
       "      <th>Property_Area</th>\n",
       "      <th>Loan_Status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LP001002</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5849</td>\n",
       "      <td>0.0</td>\n",
       "      <td>120.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LP001003</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4583</td>\n",
       "      <td>1508.0</td>\n",
       "      <td>128.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LP001005</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>LP001006</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2583</td>\n",
       "      <td>2358.0</td>\n",
       "      <td>120.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LP001008</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>6000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>141.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Loan_ID  Gender  Married  Dependents  Education  Self_Employed  ApplicantIncome  \\\n",
       "0  LP001002       1        0           0          0              0             5849   \n",
       "1  LP001003       1        1           1          0              0             4583   \n",
       "2  LP001005       1        1           0          0              1             3000   \n",
       "3  LP001006       1        1           0          1              0             2583   \n",
       "4  LP001008       1        0           0          0              0             6000   \n",
       "\n",
       "   CoapplicantIncome  LoanAmount  Loan_Amount_Term  Credit_History  Property_Area  Loan_Status  \n",
       "0                0.0       120.0             360.0             1.0              2            1  \n",
       "1             1508.0       128.0             360.0             1.0              0            0  \n",
       "2                0.0        66.0             360.0             1.0              2            1  \n",
       "3             2358.0       120.0             360.0             1.0              2            1  \n",
       "4                0.0       141.0             360.0             1.0              2            1  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "set_option('display.width', 100)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Through the output, we can corroborate columns are numeric and some of them have different scales, so it is possible we need to transform them later on.\n",
    "\n",
    "Next, we will summarize the distribution of each attribute."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>Gender</th>\n",
       "      <th>Married</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>Education</th>\n",
       "      <th>Self_Employed</th>\n",
       "      <th>ApplicantIncome</th>\n",
       "      <th>CoapplicantIncome</th>\n",
       "      <th>LoanAmount</th>\n",
       "      <th>Loan_Amount_Term</th>\n",
       "      <th>Credit_History</th>\n",
       "      <th>Property_Area</th>\n",
       "      <th>Loan_Status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "      <td>614.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.818</td>\n",
       "      <td>0.653</td>\n",
       "      <td>0.744</td>\n",
       "      <td>0.218</td>\n",
       "      <td>0.134</td>\n",
       "      <td>5403.459</td>\n",
       "      <td>1621.246</td>\n",
       "      <td>145.466</td>\n",
       "      <td>342.410</td>\n",
       "      <td>0.855</td>\n",
       "      <td>1.037</td>\n",
       "      <td>0.687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.386</td>\n",
       "      <td>0.476</td>\n",
       "      <td>1.010</td>\n",
       "      <td>0.413</td>\n",
       "      <td>0.340</td>\n",
       "      <td>6109.042</td>\n",
       "      <td>2926.248</td>\n",
       "      <td>84.181</td>\n",
       "      <td>64.429</td>\n",
       "      <td>0.352</td>\n",
       "      <td>0.787</td>\n",
       "      <td>0.464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>150.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>9.000</td>\n",
       "      <td>12.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2877.500</td>\n",
       "      <td>0.000</td>\n",
       "      <td>100.250</td>\n",
       "      <td>360.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3812.500</td>\n",
       "      <td>1188.500</td>\n",
       "      <td>125.000</td>\n",
       "      <td>360.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>5795.000</td>\n",
       "      <td>2297.250</td>\n",
       "      <td>164.750</td>\n",
       "      <td>360.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>3.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>81000.000</td>\n",
       "      <td>41667.000</td>\n",
       "      <td>700.000</td>\n",
       "      <td>480.000</td>\n",
       "      <td>1.000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Gender  Married  Dependents  Education  Self_Employed  ApplicantIncome  CoapplicantIncome  \\\n",
       "count  614.000  614.000     614.000    614.000        614.000          614.000            614.000   \n",
       "mean     0.818    0.653       0.744      0.218          0.134         5403.459           1621.246   \n",
       "std      0.386    0.476       1.010      0.413          0.340         6109.042           2926.248   \n",
       "min      0.000    0.000       0.000      0.000          0.000          150.000              0.000   \n",
       "25%      1.000    0.000       0.000      0.000          0.000         2877.500              0.000   \n",
       "50%      1.000    1.000       0.000      0.000          0.000         3812.500           1188.500   \n",
       "75%      1.000    1.000       1.000      0.000          0.000         5795.000           2297.250   \n",
       "max      1.000    1.000       3.000      1.000          1.000        81000.000          41667.000   \n",
       "\n",
       "       LoanAmount  Loan_Amount_Term  Credit_History  Property_Area  Loan_Status  \n",
       "count     614.000           614.000         614.000        614.000      614.000  \n",
       "mean      145.466           342.410           0.855          1.037        0.687  \n",
       "std        84.181            64.429           0.352          0.787        0.464  \n",
       "min         9.000            12.000           0.000          0.000        0.000  \n",
       "25%       100.250           360.000           1.000          0.000        0.000  \n",
       "50%       125.000           360.000           1.000          1.000        1.000  \n",
       "75%       164.750           360.000           1.000          2.000        1.000  \n",
       "max       700.000           480.000           1.000          2.000        1.000  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "set_option('precision', 3)\n",
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can observe our dataset has the same range, but they differ in mean values, so Standardization could be beneficial.\n",
    "\n",
    "Let's see the class distribution from the Loan_Status"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Loan_Status\n",
       "0    192\n",
       "1    422\n",
       "dtype: int64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(['Loan_Status']).size()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The class `Loans_Status` is imbalanced between YES(1) and NO(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5.2 Unimodal Data Visualizations\n",
    "\n",
    "It is useful to look at our data through different visualization techniques in order to get insights about its distribution.\n",
    "\n",
    "Let's plot a histogram to observe the dataset class distribution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x648 with 12 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.hist(sharex=False, sharey=False, xlabelsize=1, ylabelsize=1, figsize=(9,9))\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see most of the class distributions are bimodal. For other attributes, the distribution is skewed right or skewed left and Gaussian-like.\n",
    "\n",
    "We can observe easily this kind of distribution through [Density Plots](https://www.data-to-viz.com/graph/density.html#:~:text=A%20density%20plot%20is%20a,used%20in%20the%20same%20concept)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x864 with 12 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.plot(kind='density', subplots=True, layout=(4,3), sharex=False, legend=False, fontsize=1, figsize=(10,12))\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's look at the spread of attribute values using [whisker plots](https://en.wikipedia.org/wiki/Box_plot).  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "boxplot = df.boxplot(column=['Gender', 'Married', 'Dependents', 'Education', 'Self_Employed',\n",
    "       'ApplicantIncome', 'CoapplicantIncome', 'LoanAmount', 'Loan_Amount_Term', 'Credit_History',\n",
    "       'Property_Area', 'Loan_Status'], figsize=(13, 9), rot=50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Except for `ApplicantIncome`(Applicant income) and `CoapplicantIncome`(Coapplicant income), the attributes are not spread. When modeling it is possible to get some benefits if we standardize in order to line up mean values."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6 Feature Selection\n",
    "\n",
    "\n",
    "To shorten time computation and the performance of the Machine Learning Model we often prune the number of input features [2].\n",
    "\n",
    "In this section, we will visualize the correlation between features through a [Heatmap](https://en.wikipedia.org/wiki/Heat_map).  A positive correlation means an increment in one value of a feature will increment the value of the target feature. On the other hand, a negative correlation means an increase in one value of a feature will decrease the value target feature [2].\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "\n",
    "X = df.iloc[:,1:12]  #independent column\n",
    "y = df.iloc[:,12]    #target column\n",
    "\n",
    "# get correlations of each features in dataset\n",
    "corrmat = df.corr(method='pearson')\n",
    "top_corr_features = corrmat.index\n",
    "plt.figure(figsize=(20,20))\n",
    "\n",
    "# plot heat map\n",
    "g=sns.heatmap(df[top_corr_features].corr(),annot=True,cmap=\"RdYlGn\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "According to the output, the correlation between features is not significant, we will thus use all features to train the model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6.1 Validation Dataset\n",
    "\n",
    "Next, we will select a hold-out set. This dataset will be used to evaluate the performance of our model. We will use 70 % of our dataset for modeling and 30 % for validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# replace values of the target features\n",
    "df['Loan_Status'] = df.Loan_Status.replace({0: 'N', 1: 'Y'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# split-out validation dataset\n",
    "array = df.values\n",
    "X = array[:, 1:12].astype(float)\n",
    "Y = array[:, -1]\n",
    "validation_size = 0.30\n",
    "seed = 7\n",
    "X_train, X_validation, Y_train, Y_validation = train_test_split(X, Y, test_size=validation_size, random_state=seed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7 Evaluating Algorithms: Baseline\n",
    "\n",
    "To evaluate Machine Learning algorithms and select the best one, we will design a test harness with 10-fold cross-validation, and accuracy as a performance metric"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# test options and evaluation metric\n",
    "num_folds = 10\n",
    "seed = 7\n",
    "scoring = 'accuracy'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will create a baseline in order to evaluate the performance of a set of algorithms capable of solving our problem.\n",
    "\n",
    "The set of algorithms is as follows:\n",
    "\n",
    "- ***Linear Algorithms:*** Logistic Regression (LR).\n",
    "- ***Nonlinear Algorithms:*** Classification and Regression Trees (CART), Support Vector\n",
    "Machines (SVM), Gaussian Naive Bayes (NB), and k-Nearest Neighbors (KNN)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# spot-check Algoriths\n",
    "models = []\n",
    "models.append(('LR', LogisticRegression()))\n",
    "models.append(('KNN', KNeighborsClassifier()))\n",
    "models.append(('CART', DecisionTreeClassifier()))\n",
    "models.append(('NB', GaussianNB()))\n",
    "models.append(('SVM', SVC()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LR: 0.808915 (0.032245)\n",
      "KNN: 0.631728 (0.059356)\n",
      "CART: 0.671318 (0.045887)\n",
      "NB: 0.787929 (0.023846)\n",
      "SVM: 0.692193 (0.070921)\n"
     ]
    }
   ],
   "source": [
    "results = []\n",
    "names = []\n",
    "for name, model in models:\n",
    "    kfold = KFold(n_splits=num_folds, random_state=seed)\n",
    "    cv_results = cross_val_score(model, X_train, Y_train, cv=kfold, scoring=scoring)\n",
    "    results.append(cv_results)\n",
    "    names.append(name)\n",
    "    msg = \"%s: %f (%f)\" % (name, cv_results.mean(), cv_results.std())\n",
    "    print(msg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The above code snippet uses the default parameters of the selected algorithm and evaluates each algorithm based on its mean and standard deviation. According to the output, it is worth further studying `SVM` and `LR`.\n",
    "\n",
    "Next, we will use Whisker plots to visualize the distribution of accuracy values calculated above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# compare algorithms \n",
    "fig = plt.figure(figsize=(9,9))\n",
    "fig.suptitle('Algorithm Comparison')\n",
    "ax = fig.add_subplot(111)\n",
    "plt.boxplot(results)\n",
    "ax.set_xticklabels(names)\n",
    "ax.set_ylabel('Accuracy')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The results show an almost similar distribution between `SVM` and `LR`, we can see the algorithm with less variance is `NB`, though."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 8 Evaluating Algorithms: Standardize Data\n",
    "\n",
    "We think varying distributions of the dataset are affecting the algorithm performance. So, we need to evaluate each algorithm based on a  [standardized](https://en.wikipedia.org/wiki/Feature_scaling) copy (SC) of the dataset. SC means each attribute has a mean of 0 and SD of 1. To do that we use pipelines that standardize the dataset, build the model for each fold in the [cross validation test](https://towardsdatascience.com/why-and-how-to-do-cross-validation-for-machine-learning-d5bd7e60c189). This procedure allows us to prevent data leakage and carry out a better estimation of the algorithm's performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ScaledLR: 0.811240 (0.034929)\n",
      "ScaledKNN: 0.780731 (0.034462)\n",
      "ScaledCART: 0.687542 (0.032606)\n",
      "ScaledNB: 0.794906 (0.024752)\n",
      "ScaledSVM: 0.792525 (0.038202)\n"
     ]
    }
   ],
   "source": [
    "# standardize the dataset\n",
    "pipelines = []\n",
    "pipelines.append(('ScaledLR', Pipeline([('Scaler', StandardScaler()), ('LR', LogisticRegression())])))\n",
    "pipelines.append(('ScaledKNN', Pipeline([('Scaler', StandardScaler()), ('KNN', KNeighborsClassifier())])))\n",
    "pipelines.append(('ScaledCART', Pipeline([('Scaler', StandardScaler()), ('CART', DecisionTreeClassifier())])))\n",
    "pipelines.append(('ScaledNB', Pipeline([('Scaler', StandardScaler()), ('NB', GaussianNB())])))\n",
    "pipelines.append(('ScaledSVM', Pipeline([('Scaler', StandardScaler()), ('SVM', SVC())])))\n",
    "\n",
    "results = []\n",
    "names = []\n",
    "\n",
    "for name, model in pipelines:\n",
    "    kfold = KFold(n_splits=num_folds, random_state=seed)\n",
    "    cv_results = cross_val_score(model, X_train, Y_train, cv=kfold, scoring=scoring)\n",
    "    results.append(cv_results)\n",
    "    names.append(name)\n",
    "    msg = \"%s: %f (%f)\" % (name, cv_results.mean(), cv_results.std())\n",
    "    print(msg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The standardized method also points out that `SVM` and `LR` have the best performance. Next, we will plot the distributions of accuracy scores through box plots."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# compare algorithms\n",
    "fig = plt.figure(figsize=(9,9))\n",
    "fig.suptitle(\"Scaled Algoritm Comparison\")\n",
    "ax = fig.add_subplot(111)\n",
    "plt.boxplot(results)\n",
    "ax.set_xticklabels(names)\n",
    "ax.set_ylabel('Accuracy')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 9 Algorithm Tuning\n",
    "\n",
    "In this section, we will delve deeper into the parameters of `SVM` and `Logistic Regression` algorithms to enhance their performance.\n",
    "\n",
    "### 9.1 Tuning SVM\n",
    "\n",
    "We will begin by tunning the value of the regularization parameter `C`.  This parameter hinders learning complex models so that we can avoid [overfitting](https://en.wikipedia.org/wiki/Overfitting) [4]. The higher the values of `C` the lower the regularization [5]. In other words, if the `C` parameter is high we end up adjusting our model to all point features, so our model is likely overfitting. We will also tune the Kernel function parameter, so the best Kernel parameter will have a high-class separability [6]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score: 0.808915 using {'C': 0.01, 'kernel': 'linear'}\n",
      "Mean Test Score: 0.694518, STD (0.074408) with: {'C': 0.001, 'kernel': 'linear'}\n",
      "Mean Test Score: 0.696844, STD (0.071884) with: {'C': 0.001, 'kernel': 'poly'}\n",
      "Mean Test Score: 0.694518, STD (0.074408) with: {'C': 0.001, 'kernel': 'rbf'}\n",
      "Mean Test Score: 0.694518, STD (0.074408) with: {'C': 0.001, 'kernel': 'sigmoid'}\n",
      "Mean Test Score: 0.808915, STD (0.043647) with: {'C': 0.01, 'kernel': 'linear'}\n",
      "Mean Test Score: 0.696844, STD (0.071884) with: {'C': 0.01, 'kernel': 'poly'}\n",
      "Mean Test Score: 0.694518, STD (0.074408) with: {'C': 0.01, 'kernel': 'rbf'}\n",
      "Mean Test Score: 0.694518, STD (0.074408) with: {'C': 0.01, 'kernel': 'sigmoid'}\n",
      "Mean Test Score: 0.808915, STD (0.043647) with: {'C': 0.1, 'kernel': 'linear'}\n",
      "Mean Test Score: 0.764507, STD (0.048467) with: {'C': 0.1, 'kernel': 'poly'}\n",
      "Mean Test Score: 0.745847, STD (0.059565) with: {'C': 0.1, 'kernel': 'rbf'}\n",
      "Mean Test Score: 0.808915, STD (0.043647) with: {'C': 0.1, 'kernel': 'sigmoid'}\n",
      "Mean Test Score: 0.808915, STD (0.043647) with: {'C': 1, 'kernel': 'linear'}\n",
      "Mean Test Score: 0.797176, STD (0.040461) with: {'C': 1, 'kernel': 'poly'}\n",
      "Mean Test Score: 0.792525, STD (0.038202) with: {'C': 1, 'kernel': 'rbf'}\n",
      "Mean Test Score: 0.794906, STD (0.041162) with: {'C': 1, 'kernel': 'sigmoid'}\n",
      "Mean Test Score: 0.808915, STD (0.043647) with: {'C': 10, 'kernel': 'linear'}\n",
      "Mean Test Score: 0.752879, STD (0.033436) with: {'C': 10, 'kernel': 'poly'}\n",
      "Mean Test Score: 0.757641, STD (0.032758) with: {'C': 10, 'kernel': 'rbf'}\n",
      "Mean Test Score: 0.697010, STD (0.053792) with: {'C': 10, 'kernel': 'sigmoid'}\n"
     ]
    }
   ],
   "source": [
    "# tune scaled SVM\n",
    "scaler = StandardScaler().fit(X_train)\n",
    "rescaledX = scaler.transform(X_train)\n",
    "c_values = [0.001, 0.01, 0.1, 1, 10]\n",
    "kernel_values = ['linear', 'poly', 'rbf', 'sigmoid']\n",
    "param_grid = dict(C=c_values, kernel=kernel_values)\n",
    "model = SVC()\n",
    "kfold = KFold(n_splits=num_folds, random_state=seed)\n",
    "grid = GridSearchCV(estimator=model, param_grid=param_grid, scoring=scoring, cv=kfold)\n",
    "grid_result = grid.fit(rescaledX, Y_train)\n",
    "print(\"Best score: %f using %s\" % (grid_result.best_score_, grid_result.best_params_))\n",
    "means = grid_result.cv_results_['mean_test_score']\n",
    "stds = grid_result.cv_results_['std_test_score']\n",
    "params = grid_result.cv_results_['params']\n",
    "for mean, stdev, param in zip(means, stds, params):\n",
    "    print(\"Mean Test Score: %f, STD (%f) with: %r\" % (mean, stdev, param))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "According to the output, the best performance for the `SVC` has an accuracy of ***0.808915*** with a `C` parameter of **0.01** and a ***Linear*** Kernel."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 9.2 Tuning Logistic Regression\n",
    "\n",
    "Let's apply the same procedure to the Logistic Regression algorithm. But, in this case,  we will only tune the value of the regularization parameter `C`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score: 0.811240 using {'C': 0.1}\n",
      "Mean Test Score: 0.694518, STD(0.074408) with: {'C': 0.001}\n",
      "Mean Test Score: 0.808915, STD(0.043647) with: {'C': 0.01}\n",
      "Mean Test Score: 0.811240, STD(0.034929) with: {'C': 0.1}\n",
      "Mean Test Score: 0.811240, STD(0.034929) with: {'C': 1}\n",
      "Mean Test Score: 0.811240, STD(0.034929) with: {'C': 10}\n"
     ]
    }
   ],
   "source": [
    "# tune scaled Logistic Regression\n",
    "scaler = StandardScaler().fit(X_train)\n",
    "rescaledX = scaler.transform(X_train)\n",
    "c_values = [0.001, 0.01, 0.1, 1, 10]\n",
    "param_grid = dict(C=c_values, )\n",
    "model = LogisticRegression(penalty='l2')\n",
    "kfold = KFold(n_splits=num_folds, random_state=seed)\n",
    "grid = GridSearchCV(estimator=model, param_grid=param_grid, scoring=scoring, cv=kfold)\n",
    "grid_result = grid.fit(rescaledX, Y_train)\n",
    "print(\"Best score: %f using %s\" % (grid_result.best_score_, grid_result.best_params_))\n",
    "means = grid_result.cv_results_['mean_test_score']\n",
    "stds = grid_result.cv_results_['std_test_score']\n",
    "params = grid_result.cv_results_['params']\n",
    "for mean, stdev, param in zip(means, stds, params):\n",
    "    print(\"Mean Test Score: %f, STD(%f) with: %r\" % (mean, stdev, param))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "According to the output, the best performance for the `LogisticRegression` has an accuracy of **0.811240** with a `C` parameter of **0.1**. So we opt to use the **Logistic Regression** algorithm to build our prediction model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 10 Ensemble Methods\n",
    "\n",
    "In this section we will combine various models in order to obtain the best predictive model, this technique is called **Ensemble Methods** [7]. \n",
    "\n",
    "We will test two [Boosting Methods](https://en.wikipedia.org/wiki/Boosting_(machine_learning)) and two [Bagging methods](https://en.wikipedia.org/wiki/Bootstrap_aggregating#:~:text=Bootstrap%20aggregating%2C%20also%20called%20bagging,and%20helps%20to%20avoid%20overfitting):\n",
    "\n",
    "- Boosting Methods: AdaBoost (AB) and Gradient Boosting (GBM).\n",
    "- Bagging Methods: Random Forests (RF) and Extra Trees (ET).\n",
    "\n",
    "The same test schema will be used, 10-fold cross-validation. We will not standardize data because the ensemble methods we will use are based on decision trees,  that are less prone to be affected by the data distribution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy AB: 0.773920,  STD(0.045239)\n",
      "Accuracy GBM: 0.778571,  STD(0.039141)\n",
      "Accuracy RF: 0.778571,  STD(0.027836)\n",
      "Accuracy ET: 0.743632,  STD(0.017364)\n"
     ]
    }
   ],
   "source": [
    "# ensembles\n",
    "ensembles = []\n",
    "ensembles.append(('AB', AdaBoostClassifier()))\n",
    "ensembles.append(('GBM', GradientBoostingClassifier()))\n",
    "ensembles.append(('RF', RandomForestClassifier()))\n",
    "ensembles.append(('ET', ExtraTreesClassifier()))\n",
    "results = []\n",
    "names = []\n",
    "for name, model in ensembles:\n",
    "    kfold = KFold(n_splits=num_folds, random_state=seed)\n",
    "    cv_results = cross_val_score(model, X_train, Y_train, cv=kfold, scoring=scoring)\n",
    "    results.append(cv_results)\n",
    "    names.append(name)\n",
    "    msg = \"Accuracy %s: %f,  STD(%f)\" % (name, cv_results.mean(), cv_results.std())\n",
    "    print(msg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's  plot the distributions of accuracy scores through box plots."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(9,9))\n",
    "fig.suptitle('Ensemble Algorith Comparison')\n",
    "ax = fig.add_subplot(111)\n",
    "plt.boxplot(results)\n",
    "ax.set_xticklabels(names)\n",
    "ax.set_ylabel('Accuracy')\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "According to the outputs, the performance of ensemble methods is no better than Linear Algorithms, so we will finalize our model with the ***Logistic Regression*** algorithm."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 11 Finalizing  Model\n",
    "\n",
    "Our findings show that the **Logistic Regression** model is less complex and more stable than the other tested models. In this section, we will train the **Logistic Regression** model for the whole dataset and make predictions for the validation dataset to see if our findings are also valid for the validation dataset. As before, this validation will be carried out on a standardized version of the test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score: 0.81\n"
     ]
    }
   ],
   "source": [
    "# prepare the model\n",
    "scaler = StandardScaler().fit(X_train)\n",
    "rescaledX = scaler.transform(X_train)\n",
    "model = LogisticRegression(C= 0.1, penalty='l2')\n",
    "model.fit(rescaledX, Y_train)\n",
    "\n",
    "# estimate accuarcy on validation dataset\n",
    "rescaledValidationX = scaler.transform(X_validation)\n",
    "predictions = model.predict(rescaledValidationX)\n",
    "print('Accuracy Score:',round(accuracy_score(Y_validation, predictions), 2));"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that **81%** of our predictions were correct. Now, let's create a table containing the classification report.\n",
    "\n",
    "Next, we will create a data frame containing the classification report."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "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>precision</th>\n",
       "      <th>recall</th>\n",
       "      <th>f1-score</th>\n",
       "      <th>support</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>N</th>\n",
       "      <td>0.933</td>\n",
       "      <td>0.459</td>\n",
       "      <td>0.615</td>\n",
       "      <td>61.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Y</th>\n",
       "      <td>0.787</td>\n",
       "      <td>0.984</td>\n",
       "      <td>0.875</td>\n",
       "      <td>124.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   precision  recall  f1-score  support\n",
       "N      0.933   0.459     0.615     61.0\n",
       "Y      0.787   0.984     0.875    124.0"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# generate classification report\n",
    "cr = classification_report(Y_validation, predictions, output_dict=True)\n",
    "dfcr = pd.DataFrame.from_dict(cr)\n",
    "dfcr = dfcr[['N', 'Y']].T\n",
    "dfcr = dfcr[['precision', 'recall', 'f1-score', 'support']]\n",
    "dfcr"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "According to the output, `f1-score` and `support` values indicate that our dataset is imbalanced. So, our `Accuracy Score` (81%) is not completely accurate because this metrics is reliable when we have a balanced dataset (similar class distribution). This result indicates that our datasets could need stratified sampling or rebalancing [8] "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 12 Predicting on test dataset\n",
    "\n",
    "In this section, we will use our trained model to predict loan status on the test dataset.\n",
    "\n",
    "First we will load and visualize the test dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>Loan_ID</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Married</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>Education</th>\n",
       "      <th>Self_Employed</th>\n",
       "      <th>ApplicantIncome</th>\n",
       "      <th>CoapplicantIncome</th>\n",
       "      <th>LoanAmount</th>\n",
       "      <th>Loan_Amount_Term</th>\n",
       "      <th>Credit_History</th>\n",
       "      <th>Property_Area</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LP001015</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>No</td>\n",
       "      <td>5720</td>\n",
       "      <td>0</td>\n",
       "      <td>110.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Urban</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LP001022</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>1</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>No</td>\n",
       "      <td>3076</td>\n",
       "      <td>1500</td>\n",
       "      <td>126.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Urban</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LP001031</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>2</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>No</td>\n",
       "      <td>5000</td>\n",
       "      <td>1800</td>\n",
       "      <td>208.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Urban</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>LP001035</td>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>2</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>No</td>\n",
       "      <td>2340</td>\n",
       "      <td>2546</td>\n",
       "      <td>100.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Urban</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LP001051</td>\n",
       "      <td>Male</td>\n",
       "      <td>No</td>\n",
       "      <td>0</td>\n",
       "      <td>Not Graduate</td>\n",
       "      <td>No</td>\n",
       "      <td>3276</td>\n",
       "      <td>0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Urban</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Loan_ID Gender Married Dependents     Education Self_Employed  ApplicantIncome  \\\n",
       "0  LP001015   Male     Yes          0      Graduate            No             5720   \n",
       "1  LP001022   Male     Yes          1      Graduate            No             3076   \n",
       "2  LP001031   Male     Yes          2      Graduate            No             5000   \n",
       "3  LP001035   Male     Yes          2      Graduate            No             2340   \n",
       "4  LP001051   Male      No          0  Not Graduate            No             3276   \n",
       "\n",
       "   CoapplicantIncome  LoanAmount  Loan_Amount_Term  Credit_History Property_Area  \n",
       "0                  0       110.0             360.0             1.0         Urban  \n",
       "1               1500       126.0             360.0             1.0         Urban  \n",
       "2               1800       208.0             360.0             1.0         Urban  \n",
       "3               2546       100.0             360.0             NaN         Urban  \n",
       "4                  0        78.0             360.0             1.0         Urban  "
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# load test dataset\n",
    "df_test = pd.read_csv('test.csv')\n",
    "df_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's get a count of missing values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Loan_ID               0\n",
       "Gender               11\n",
       "Married               0\n",
       "Dependents           10\n",
       "Education             0\n",
       "Self_Employed        23\n",
       "ApplicantIncome       0\n",
       "CoapplicantIncome     0\n",
       "LoanAmount            5\n",
       "Loan_Amount_Term      6\n",
       "Credit_History       29\n",
       "Property_Area         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sum of missing values \n",
    "df_test.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following lines of code will allow us to get rid of missing values by imputing them, and change the data type of some columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "# imputing both categorical and numerical values\n",
    "imp = SimpleImputer(missing_values=np.nan, strategy='most_frequent')\n",
    "df_test = pd.DataFrame(imp.fit_transform(df_test),columns=df_test.columns,index=df_test.index)\n",
    "\n",
    "# change data types\n",
    "df_test = df_test.astype({'ApplicantIncome': np.int64, 'CoapplicantIncome':float,\n",
    "                'LoanAmount':float,'Loan_Amount_Term':float, 'Credit_History':float})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check the results,"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Loan_ID               0\n",
       "Gender                0\n",
       "Married               0\n",
       "Dependents            0\n",
       "Education             0\n",
       "Self_Employed         0\n",
       "ApplicantIncome       0\n",
       "CoapplicantIncome     0\n",
       "LoanAmount            0\n",
       "Loan_Amount_Term      0\n",
       "Credit_History        0\n",
       "Property_Area         0\n",
       "GenderLabel           0\n",
       "MarriedLabel          0\n",
       "DependentsLabel       0\n",
       "EducationLabel        0\n",
       "Self_EmployedLabel    0\n",
       "Property_AreaLabel    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sum of missing values \n",
    "df_test.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we will  transform categorical by applying the `encoding_schema()` function defined above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Gender are {0: 'Female', 1: 'Male'}\""
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Gender \n",
    "encoding_schema(df_test, 'Gender', 'GenderLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Married are {0: 'No', 1: 'Yes'}\""
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Married \n",
    "encoding_schema(df_test, 'Married', 'MarriedLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Dependents are {0: '0', 1: '1', 2: '2', 3: '3+'}\""
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Dependents\n",
    "encoding_schema(df_test, 'Dependents', 'DependentsLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Education are {0: 'Graduate', 1: 'Not Graduate'}\""
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Education\n",
    "encoding_schema(df_test, 'Education', 'EducationLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Self_Employed are {0: 'No', 1: 'Yes'}\""
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Self_Employed\n",
    "encoding_schema(df_test, 'Self_Employed', 'Self_EmployedLabel')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Labels for Property_Area are {0: 'Rural', 1: 'Semiurban', 2: 'Urban'}\""
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# apply encoding_scheme to Property_Area\n",
    "encoding_schema(df_test, 'Property_Area', 'Property_AreaLabel')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After transforming categorical data,  we will get rid of unuseful columns and rename back useful columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get rid of unuseful column\n",
    "columns_to_keep  = ['Loan_ID', 'GenderLabel', 'MarriedLabel', 'DependentsLabel', 'EducationLabel',\n",
    "       'Self_EmployedLabel', 'ApplicantIncome', 'CoapplicantIncome', 'LoanAmount',\n",
    "       'Loan_Amount_Term', 'Credit_History', 'Property_AreaLabel']\n",
    "\n",
    "df_test = df_test[columns_to_keep]\n",
    "df_test.dtypes\n",
    "\n",
    "# rename back the column names\n",
    "old_colnames  = ['GenderLabel', 'MarriedLabel', 'DependentsLabel', 'EducationLabel',\n",
    "       'Self_EmployedLabel', 'Property_AreaLabel']\n",
    "\n",
    "new_colnames  = ['Gender', 'Married', 'Dependents', 'Education',\n",
    "       'Self_Employed', 'Property_Area']\n",
    "\n",
    "col_rename_dict = {i:j for i,j in zip(old_colnames,new_colnames)}\n",
    "df_test = df_test.rename(columns=col_rename_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's visualize the the results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "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>Loan_ID</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Married</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>Education</th>\n",
       "      <th>Self_Employed</th>\n",
       "      <th>ApplicantIncome</th>\n",
       "      <th>CoapplicantIncome</th>\n",
       "      <th>LoanAmount</th>\n",
       "      <th>Loan_Amount_Term</th>\n",
       "      <th>Credit_History</th>\n",
       "      <th>Property_Area</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LP001015</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5720</td>\n",
       "      <td>0.0</td>\n",
       "      <td>110.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LP001022</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3076</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>126.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LP001031</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>1800.0</td>\n",
       "      <td>208.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>LP001035</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2340</td>\n",
       "      <td>2546.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LP001051</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3276</td>\n",
       "      <td>0.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>360.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Loan_ID  Gender  Married  Dependents  Education  Self_Employed  ApplicantIncome  \\\n",
       "0  LP001015       1        1           0          0              0             5720   \n",
       "1  LP001022       1        1           1          0              0             3076   \n",
       "2  LP001031       1        1           2          0              0             5000   \n",
       "3  LP001035       1        1           2          0              0             2340   \n",
       "4  LP001051       1        0           0          1              0             3276   \n",
       "\n",
       "   CoapplicantIncome  LoanAmount  Loan_Amount_Term  Credit_History  Property_Area  \n",
       "0                0.0       110.0             360.0             1.0              2  \n",
       "1             1500.0       126.0             360.0             1.0              2  \n",
       "2             1800.0       208.0             360.0             1.0              2  \n",
       "3             2546.0       100.0             360.0             1.0              2  \n",
       "4                0.0        78.0             360.0             1.0              2  "
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we will prepare our test dataset for predictions and use our trained model to predict loan_status."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_names = ['Gender', 'Married', 'Dependents', 'Education', 'Self_Employed',\n",
    "       'ApplicantIncome', 'CoapplicantIncome', 'LoanAmount', 'Loan_Amount_Term', 'Credit_History',\n",
    "       'Property_Area']\n",
    "\n",
    "# define index\n",
    "df_test.index = df_test['Loan_ID']\n",
    "\n",
    "# define validation features\n",
    "X_validation = df_test[feature_names]\n",
    "\n",
    "# rescale validation set\n",
    "rescaledValidationX = scaler.transform(X_validation)\n",
    "    \n",
    "# calculate the predictions for each class\n",
    "X_predict = model.predict(rescaledValidationX)\n",
    "\n",
    "# create the solution dataframe\n",
    "result = pd.DataFrame(data=X_predict, index=df_test['Loan_ID'], columns=['Loan_Status'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we will save our result as a CSV file and submit it to the Solution Checker of the [Practice Problem Loan Prediction](https://datahack.analyticsvidhya.com/contest/practice-problem-loan-prediction/). Our submission got an accuracy of **0.77** on the leaderboard."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save the solution as an csv file\n",
    "result.to_csv('solution.csv', header=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 13 Conclusion\n",
    "\n",
    "Through this notebook, we showed you a simple workflow to carry out a binary classification using Machine Learning. This is a methodology proposal that is open to be improved. Although we got moderate accuracy (0.81 and 0.77), future works are needed to deal with imbalanced data and enhance accuracy. If you have a comment or suggestion please hit the Contact tab of my web page (https://acoiman.github.io/#contact) and send us a message."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 14 References"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[1]: https://machinelearningmastery.com/machine-learning-with-python/ \"Machine Learning Mastery With Python\".\n",
    "\n",
    "\n",
    "[2]: https://machinelearningmastery.com/feature-selection-with-real-and-categorical-data/ \"How to Choose a Feature Selection Method For Machine Learning\".\n",
    "\n",
    "[3]: https://towardsdatascience.com/feature-selection-techniques-in-machine-learning-with-python-f24e7da3f36e \"Feature Selection Techniques in Machine Learning with Python\".\n",
    "\n",
    "[4]:https://towardsdatascience.com/regularization-in-machine-learning-76441ddcf99a \"Regularization in Machine Learning\".\n",
    "\n",
    "[5]:  Müller, A. Guido, S. (2016). Introduction to machine learning with python. O'Reilly Media, Incorporated. USA.\n",
    "\n",
    "[6]: Liu, Z., & Xu, H. (2014). Kernel parameter selection for support vector machine classification. Journal of Algorithms & Computational Technology, 8(2), 163-177.\n",
    "\n",
    "[7]: https://towardsdatascience.com/ensemble-methods-in-machine-learning-what-are-they-and-why-use-them-68ec3f9fef5f#:~:text=Ensemble%20methods%20is%20a%20machine,machine%20learning%20and%20model%20building. Ensemble Methods in Machine Learning: What are They and Why Use Them?\n",
    "\n",
    "[8]: https://www.scikit-yb.org/en/latest/api/classifier/classification_report.html?highlight=Classification%20Report. Classification Report.\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.8"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}