Discriminative v.s.Generative

For example, to understand ：

Suppose we have two kinds of animals , Elephant ( y = 1) And dogs ( y = 0).x It's the eigenvector of the animal .

Given a training set , Logistic regression or perceptron algorithm （ Basically ） Trying to find a straight line —— That's the decision boundary —— Separate the elephant from the dog . then , To classify new animals as elephants or dogs , It checks which side of the decision boundary it falls on , And make corresponding predictions . We call these Discriminant learning algorithm .

This is a different way . First , Look at the elephant , We can build a model of what an elephant looks like . then , Look at the dog , We can build a separate model to understand what a dog looks like . Last , In order to classify new animals , We can match the new animal to the elephant model , And match it with the dog model , See if the new animal looks more like an elephant or more like the dog we saw in the training set .  We call these Generative learning algorithm .

Definition understanding ：

More formally, given a set of data instances X and a set of labels Y:

• Generative models capture the joint probability p(X, Y), or just p(X) if there are no labels.
• Discriminative models capture the conditional probability p(Y | X).

A generative model includes the distribution of the data itself, and tells you how likely a given example is. For example, models that predict the next word in a sequence are typically generative models (usually much simpler than GANs) because they can assign a probability to a sequence of words.

A discriminative model ignores the question of whether a given instance is likely, and just tells you how likely a label is to apply to the instance.

Note that this is a very general definition. There are many kinds of generative model. GANs are just one kind of generative model.