BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:IC Monday Seminar : Learning to Behave by Reading
DTSTART:20111010T161500
DTSTAMP:20260509T211801Z
UID:9463a6bd0b7d77a9b022c0b0093fe753f5bbd10153f3b27ecf18f912
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Regina Barzilay\, MIT - Invited by Prof. Christoph Koch\
 nAbstract In this talk I will address the problem of grounding linguistic 
 analysis in control applications such as game playing and robot navigation
 . We assume access to natural language documents that describe the desired
  behavior of a control algorithm (e.g. game strategy guides). Our goal is 
 to demonstrate that knowledge automatically extracted from such documents 
 can dramatically improve performance of the target application. First I wi
 ll present a reinforcement learning algorithm for learning to map natural 
 language instructions to executable actions. This technique has enabled au
 tomation of tasks that until now have required human participation --- for
  example automatically configuring software by consulting how-to guides. N
 ext I will present a Monte-Carlo search algorithm for game playing that in
 corporates information from game strategy guides. In this framework the ta
 sk of text interpretation is formulated as a probabilistic model that is t
 rained based on feedback from Monte-Carlo search. When applied to the Civi
 lization strategy game a language-empowered player outperforms its traditi
 onal counterpart by a significant margin. 
LOCATION:INM 202
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR