BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Game Playing Meets Game Theory: Strategic Learning from Simulated 
 Play - wellman@umich.edu
DTSTART:20200305T110000Z
DTEND:20200305T120000Z
UID:TALK140311@talks.cam.ac.uk
CONTACT:Adrian Weller
DESCRIPTION:Recent breakthroughs in AI game-playing -- AlphaGo (Go)\, Alph
 aZero (Chess\, Shogi +)\, AlphaStar (StarCraft II)\, Libratus and DeepStac
 k (Poker) -- have demonstrated superhuman performance in a range of recrea
 tional strategy games. Extending beyond artificial domains presents severa
 l challenges\, but the basic idea of learning from simulated play employed
  in most of these systems is broadly applicable to any domain that can be 
 accurately simulated. This thread of work naturally dovetails with methods
  developed in the Strategic Reasoning Group at Michigan for reasoning abou
 t simulation-based games. I will recap some of this work\, with emphasis o
 n how new advances in deep reinforcement learning can contribute to a majo
 r broadening of the scope of game-theoretic reasoning for complex multiage
 nt domains.
LOCATION:Engineering Department\, CBL Room BE-438.
END:VEVENT
END:VCALENDAR