BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Improving Generative Adversarial Networks using Game Theory and St
 atistics - Constantinos Daskalakis (MIT)
DTSTART:20180329T140000Z
DTEND:20180329T150000Z
UID:TALK103277@talks.cam.ac.uk
CONTACT:Carola-Bibiane Schoenlieb
DESCRIPTION:Generative Adversarial Networks\, a.k.a. GANs\, have found gre
 at applicability in machine learning in applications ranging from image sy
 nthesis to compressed sensing\, domain adaptation and super-resolution. Th
 ey are defined by setting up a two-player zero-sum game between two neural
  networks\, trained using gradient descent on samples from a target distri
 bution. Despite their practical success\, GANs pose great challenges for b
 oth optimization and statistics. Their training suffers from oscillations\
 , and they are difficult to scale to high-dimensional settings. We study h
 ow game-theoretic and statistical techniques can be brought to bare on the
 se important challenges.\n\n\n\nBio: Constantinos Daskalakis is a professo
 r of computer science and electrical engineering at MIT. He holds a diplom
 a in electrical and computer engineering from the National Technical Unive
 rsity of Athens\, and a Ph.D. in electrical engineering and computer scien
 ces from UC-Berkeley. His research interests lie in theoretical computer s
 cience and its interface with economics\, game theory\, probability\, lear
 ning 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 Pri
 ze\, the 2011 Ruth and Joel Spira Award for Distinguished Teaching\, the 2
 012 Microsoft Research Faculty Fellowship\, and the 2015 Research and Deve
 lopment Award by the Vatican Giuseppe Sciacca Foundation. He is also a rec
 ipient of Best Paper awards at the ACM Conference on Economics and Computa
 tion in 2006 and in 2013.
LOCATION:MR14\, Centre for Mathematical Sciences
END:VEVENT
END:VCALENDAR