Logistic regression is a means of fitting a line to data that fits into a binary classification (class). More specifically, $y$ can only be one of two values. Usually, these two values are represented as true (1) or false (0), and they are called the positive class and negative class, respectively.

Even though the algorithm is called logistic regression, it is actually a classification problem and is only called this because of historical purposes.

1_dm6ZaX5fuSmuVvM4Ds-vcg.jpg

Defining the Model

The Decision Boundary

The Cost Function

Gradient Descent for Logistic Regression

Regularized Logistic Regression