TITLE: Trajectory Trees: Combining State and Temporal Abstraction for Markov Decision Problems

PRESENTER: Will Uther

AFFILIATION: Department of Computer Science, Carnegie Mellon University, Pittsburgh, PA.

DATE: Tuesday, 8th January 2002

TIME: 12:00noon - 1.00pm

PLACE: CSE K17 1st Floor Seminar Room

ABSTRACT:

There are two types of abstraction defined in the Markov Decision Problem (MDP) literature, state and temporal abstraction. I will present a new algorithm, the Trajectory Tree algorithm, that can automatically generate state abstractions, and can make use of supplied temporal abstractions.

BIOGRAPHY OF SPEAKER:

William Uther is a doctoral student in the computer science department of Carnegie Mellon University working with Manuela Veloso. William graduated from Sydney University in 1994.

Seminar Convenor:

Bernhard Hengst
Tel: +61-2-9385-3988,
E-mail: bernhardh@cse.unsw.edu.au
School of Computer Science & Engineering, UNSW.