Logistic regression is a classification algorithm used for binary classification. It uses a logistic function to formulate the interaction between the input features and the target variable. It is one of the simplest classification techniques that is used to model a binary dependent variable.