Classification methods
If the response variable is qualitative/categorical (takes on categorical values such as gender, loan default, marital status, and such), then the problem can be called a classification problem regardless of the explanatory variables' type. There are various types of classification methods, but we will focus on logistic regression and Support Vector Machines in this section.
Following are a few examples of some implications of classification methods:
A customer buys a product or does not buy it
A person is diabetic or not diabetic
An individual applying for a loan would default or not
An e-mail receiver would read the e-mail or not
Logistic regression
Logistic regression measures the relation between the explanatory variables and the categorical response variable. We do not use linear regression for the categorical response variable because the response variable is not on a continuous scale and hence the error terms are not normally distributed.
So logistic regression is a...
