Marc
Lanctot's
Corner of the World-Wide Web
Overview
This web site has moved! You should be redirected in 5 seconds. If not, click here.
I am a Ph.D. candidate working under the supervision of Michael Bowling. I am generally interested in AI, machine learning, and games. I would like to know how intelligent beings make decisions, how they represent information and reason about it efficiently, and how interaction can lead to knowledge gain.
Decision-making is an inherent component of game-playing. Most games are easy to observe; each instance is easy to record and learn from. People like playing games so there certainly is a wealth of data to learn from. In addition, games are often played quite competitively; the decisions made by the players usually reflect profoundly considered thoughts and conclusions. Therefore, game-playing is a perfect context from which to study intelligent decision-making.
I am particularly interested in computational game theory and learning through self-play. These days I am working on regret minimization methods for solving large extensive-form games. While my work focuses on imperfect information games, it is not tied to a specific game; I am interested in the behavior of these algorithms in the general case, as a function of the type of imperfect information, or as a function of the structure of the game. In the past year, I have also become interested in Monte-Carlo Tree Search and other Monte-Carlo search algorithms.
If you would like to know more feel free to contact me about anything, either by email or dropping by the AI lab.
Schedule
I am now living in Toronto and so will be commuting to Edmonton every second Thursday. I normally arrive on campus at 12-12:30. If I stay overnight, I leave some time Friday afternoon. Otherwise, I get the last flight on Thursday evening. To catch the latest flight I need to leave campus by 4:15 to get the last Sky Shuttle at 4:30.Last Modified: Apr 12th, 2012
msfmustdie@hotmail.com - w8ting-room - v2005