The Hong Kong University of Science and Technology
Department of Computer Science


PhD Qualifying examination


"Algorithms for Partially Observable Markov Decision Processes"


By Mr. Weihong Zhang


Abstract:

Partially Observable Markov Decision Processes(POMDPs) provide an elegant 
framework for a wide range of sequential decision making problems. In this survey,
I present an overview on algorithms for solving POMDPs. We start from introducing
the background knowledge and model description. Afterwards, the problem is
defined and the difficulties in solving the problem are identified. We then survey
three classes of existing algorithms: value iteration algorithms, policy
iteration algorithms and machine learning algorithms.



Date:                   Friday, 23 February 2001

Time:                   3:15 p.m. to 5:15 p.m.
        
Venue:                  Room 1402
			Lift 25-26
                        

Chairman:               Dr. James T Y Kwok

Committee Members:      Dr. Nevin Zhang (Supervisor)
                        Dr. Dit-Yan Yeung
                        Dr. Mordecai Golin
			Dr. Fangzhen Lin


**** ALL are Welcome ****