BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Cooperative Inverse RL - Robert Pinsler\; Adria Garriga Alonso
DTSTART:20171026T123000Z
DTEND:20171026T140000Z
UID:TALK94210@talks.cam.ac.uk
CONTACT:Alessandro Davide Ialongo
DESCRIPTION:*Abstract:*\n\nThe value alignment problem consists in ensurin
 g the values of an AI system align with the values of its operator. A pote
 ntial solution to this problem is formalised as the Inverse Reinforcement 
 Learning (IRL) setting. In IRL\, the goal is to infer the reward function 
 of an agent (a human)\, just from observing its\nbehaviour in the environm
 ent. In Cooperative IRL\, the agents are allowed to interact. From this\, 
 more effective teaching strategies than passive observation\nemerge. We wi
 ll talk about formalising this problem\, and an algorithm to approximate g
 ood teaching strategies.\n\n*Recommended reading:* \n\nMain Paper: \n* Coo
 perative Inverse Reinforcement Learning: https://papers.nips.cc/paper/6420
 -cooperative-inverse-reinforcement-learning.pdf\n\nOptional background rea
 ding on Inverse Reinforcement Learning:\n* Apprenticeship Learning via Inv
 erse Reinforcement Learning: http://ai.stanford.edu/~ang/papers/icml04-app
 rentice.pdf\n* Maximum Entropy Inverse Reinforcement Learning: https://www
 .aaai.org/Papers/AAAI/2008/AAAI08-227.pdf
LOCATION:Engineering Department\, CBL Seminar Room 4-38
END:VEVENT
END:VCALENDAR