BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:General Reinforcement Learning - Jan Leike (Australian National Un
 iversity)
DTSTART:20151216T110000Z
DTEND:20151216T120000Z
UID:TALK62605@talks.cam.ac.uk
CONTACT:Adrian Weller
DESCRIPTION:Reinforcement learning problems are often phrased in terms of 
 Markov decision processes (MDPs). In this talk\, we go beyond MDPs and con
 sider reinforcement learning in environments that are non-Markovian\, non-
 ergodic and only partially observable. Our focus will not be on practical 
 algorithms\, but rather on the fundamental underlying problems. How do we 
 balance exploration and exploitation? How do we explore optimally? When is
  an agent optimal? We introduce the Bayesian agent AIXI\, point out some o
 f its problems\, and discuss potential solutions.\n\nSpeaker:\nJan Leike i
 s a PhD student at the Australian National University working with Marcus 
 Hutter.
LOCATION:Engineering Department\, CBL Room BE-438
END:VEVENT
END:VCALENDAR