Implementing logistic regression with Python

We have understood the mathematics that goes behind the logistic regression algorithm. Now, let's take one dataset and implement a logistic regression model from scratch. The dataset we will be working with is from the marketing department of a bank and has data about whether the customers subscribed to a term deposit, given some information about the customer and how the bank has engaged and reached out to the customers to sell the term deposit.

Let us import the dataset and start exploring it:

import pandas as pd
bank=pd.read_csv('E:/Personal/Learning/Predictive Modeling Book/Book Datasets/Logistic Regression/bank.csv',sep=';')
bank.head()

The dataset looks as follows:

Fig. 6.6: A glimpse of the bank dataset

There are 4119 records and 21 columns. The column names are as follows:

bank.columns.values

Fig. 6.7: The columns of the bank dataset

The details of each column are mentioned in the Data Dictionary file present in the Logistic Regression folder...