Social Data Science
Generative Adversarial Networks (GANs) have proven themselves as the state-of-the-art generative models, a class of machine learning algorithms that can be used to synthesize new data. Their ability to be trained completely via backpropagation helps them become significantly easier to use in comparison to the previous generative models. Unfortunately, GANs also bring along their own baggage of problems. As a part of this session, we will explore the general idea of GANs, how to make them and where to use them. Presenter: Yash Upadhyay. Yash is a Master's student at the University of Minnesota - Twin Cities, majoring in Computer Science. His current research involves the exploration of manifold learning and generative systems in the field of computer vision. He hopes someday to build an intelligent system that can synthesize videos with accurate future predictions.
2815 Cleveland Ave