Eighteenth Annual IEEE Symposium on

LOGIC IN COMPUTER SCIENCE (LICS 2003)
Workshop on Probability in AI

June 21st, 2003, Ottawa, Canada

SITE room C0136

 

Workshop Organizer:

Doina Precup
School of Computer Science
McGill University

Email: dprecup@cs.mcgill.ca
Phone: +1 514 398-6443 
Fax: +1 514 398-3883

A lot of the recent research in artificial intelligence (AI) has focused on the role of probabilities and probabilistic reasoning for representing knowledge and reasoning about complex, unknown environments. In the LICS conferences in recent years there has been a lot of activity around the area of verification and model checking in particular, with a number of recent papers specifically focusing on probability.

There are more and more clear connections between concepts and methods used in the fields of AI and verification. For instance, although the questions asked in these two fields are different, a lot of the algorithmic approaches are based on Markov chains or Markov decision processes. There are also strong connections between notions like bisimulation and state space reduction. Given the synergy that is building up between AI and verification, it seems appropriate to have a workshop where AI researchers in probability and logic can interact with more traditional LICS participants.

Preliminary schedule

9:15-9:30
 Welcome and introductory remarks 
9:30-10:30
 Bob Givan
 Department of Electrical and Computer Engineering, Purdue University
 Approximate Policy Iteration with a Policy Language Bias
Abstract
10:30-11:00
  Coffee break
11:00-11:45
 Pascal Poupart
 Department of Computer Science,University of Toronto
 Bisimulation for Partially Observable Markov Decision Processes
Abstract
11:45-12:15
 Norm Ferns
 School of Computer Science, McGill University
 Metrics for Markov Decision Processes
Abstract
12:15-14:00
  Lunch break
14:00-15:00
  Marta Kwiatkowska
 Department of Computer Science, University of Birmingham
 Verification of randomised distributed algorithms
 Abstract
15:00-15:30
  K. Subramani
 Department of Computer Science and Electrical Engineering, West Virginia University
 On a random walk strategy for the Q2SAT problem
Abstract,Paper
15:30-16:00
  Coffee break
16:00-17:00
  Ron Parr
 Department of Computer Science, Duke University
 Linear Value Function Approximation Methods for MDPs and Markov Games
Abstract
17:00-17:15
 Concluding remarks 

 

We are happy to acknowledge the generous support of the Centre de Recherche Mathematiques of the Universite de Montreal.


Other important dates:

  • LICS Conference: June 22nd - 25th
  • Workshop: June 21st
  • Pre LICS summer school: June 2nd - 20th

Back to the LICS web page.


Last modified: 20th June 2003