What is a Generative Adversarial Network (GAN) and how does its architecture function?

A Generative Adversarial Network (GAN) consists of two neural networks: a generator and a discriminator. The generator creates synthetic data samples, while the discriminator evaluates them against real data. During training, the two networks compete in a zero-sum game: the generator aims to produce data that improves its chances of fooling the discriminator, while the discriminator strives to differentiate between real and generated samples. This adversarial process leads to the generator creating increasingly realistic data that can mimic the training distribution.