[video] Introduction to Generative Adversarial Networks (for beginners and advanced Data Scientists)

Generative Adversarial Networks are driving important new technologies in deep learning methods. With so much to learn, these two videos will help you jump into your exploration with GANs and the mathematics behind the modelling.

By Zak Jost, Research Scientist at Amazon Web Services.

Since Ian Goodfellow and team published their paper introducing Generative Adversarial Networks (GANs) in 2014, this unsupervised approach to deep learning has brought about a wave of technical advancements that essentially appear like the development of "AI creativity."

As GANs have already become an important tool to appreciate in a data scientist's repertoire of machine learning skills, the following two videos from Zak Jost will help you begin your exploration into this exciting area of deep learning.

Introduction to GANs

The first video provides a high-level overview of GANs featuring a simple illustration of the the key aspects of adversarial learning through modeling a coin-flip scenario (10 minutes).

The Math of GANs

The second video is a deeper dive into GANs with a look at the mathematics behind the modelling. Zak first explains the optimization function, then works through its solution that leads to a perfect generative model (12 minutes).


Bio: Zak Jost (@ZakJost) is Machine Learning Research Scientists at Amazon Web Services working on fraud applications. Before this, Zak built large-scale modeling tools as a Principal Data Scientist at Capital One to support the business's portfolio risk assessment efforts following a previous career as a Material Scientist in the semiconductor industry building thin-film nanomaterials.