Logistic regression is based on the logistic function or the sigmoid function. It is a type of linear model, which transforms the probabilities into discrete values using sigmoid function. So, unlike linear regression, this model can be used to map a set of values to two or more classes.
The following links have a very good explanation on logistic regression:
https://machinelearningmastery.com/logistic-regression-for-machine-learning/
https://towardsdatascience.com/logistic-regression-detailed-overview-46c4da4303bc
This project is just about experimenting logistic regression on two datasets. Both the datasets differ in number of rows and the values in risk column. In built python libraries have been used to carry out logistic regression.