Logistic regression is used for classification. The independent variable of the data is quantitative and the dependent variable of the data is binary (0 or 1) i.e. in a class or not. Instead of modeling the response (0 or 1), the logistic model’s dependent variable is the probability that a data point belongs in the class.