Discriminative model

A statistical model that does not make a general joint probability distribution like a Generative model

Instead, the posterior is modeled directly. In other words, a function/ mapping from the data to the probability of the label is determined usually based on some parametric function.

Example: Logistic regression