Improving Generative Adversarial Networks using Game Theory and Statistics Joint CCIMI-ACA seminar

With Constantinos Daskalakis (MIT)

Improving Generative Adversarial Networks using Game Theory and Statistics

Generative Adversarial Networks, a.k.a. GANs, have found great applicability in machine learning in applications ranging from image synthesis to compressed sensing, domain adaptation and super-resolution. They are defined by setting up a two-player zero-sum game between two neural networks, trained using gradient descent on samples from a target distribution. Despite their practical success, GANs pose great challenges for both optimization and statistics. Their training suffers from oscillations, and they are difficult to scale to high-dimensional settings. We study how game-theoretic and statistical techniques can be brought to bare on these important challenges.

Bio: Constantinos Daskalakis is a professor of computer science and electrical engineering at MIT . He holds a diploma in electrical and computer engineering from the National Technical University of Athens, and a Ph.D. in electrical engineering and computer sciences from UC-Berkeley. His research interests lie in theoretical computer science and its interface with economics, game theory, probability, learning and statistics. He has been honored with the 2007 Microsoft Graduate Research Fellowship, the 2008 ACM Doctoral Dissertation Award, the Game Theory and Computer Science Prize from the Game Theory Society, the 2010 Sloan Fellowship in Computer Science, the 2011 SIAM Outstanding Paper Prize, the 2011 Ruth and Joel Spira Award for Distinguished Teaching, the 2012 Microsoft Research Faculty Fellowship, and the 2015 Research and Development Award by the Vatican Giuseppe Sciacca Foundation. He is also a recipient of Best Paper awards at the ACM Conference on Economics and Computation in 2006 and in 2013.

Joint CCIMI-ACA seminar

Add to your calendar or Include in your list