We started this chapter by describing and comparing discriminative and generative models. We introduced a few probabilistic concepts, including Bayes' theorem, and described mathematically and visually the probabilistic models learned by discriminative and generative models. Next, we provided information about multiple types of generative models, including GANs, Variational Autoencoders, and reversible flow models. We mathematically derived these models, compared them to each other, showed examples of their usage, and enumerated their advantages and disadvantages.
In addition, we described the building blocks of GANs, enumerating the individual components used in the framework and examining how they can be used. Finally, we briefly exposed their strengths and limitations.
In the next chapter, you will learn how to implement your first GAN.
