Dynamic Programming Approximations for Partially Observable Stochastic Games

Author

Akshat KUMAR, Singapore Management UniversityFollow
Shlomo ZILBERSTEIN, University of Massachusetts Amherst

Publication Type

Conference Proceeding Article

Version

publishedVersion

Publication Date

5-2009

Abstract

Partially observable stochastic games (POSGs) provide a rich mathematical framework for planning under uncertainty by a group of agents. However, this modeling advantage comes with a price, namely a high computational cost. Solving POSGs optimally quickly becomes intractable after a few decision cycles. Our main contribution is to provide bounded approximation techniques, which enable us to scale POSG algorithms by several orders of magnitude. We study both the POSG model and its cooperative counterpart, DEC-POMDP. Experiments on a number of problems confirm the scalability of our approach while still providing useful policies.

Discipline

Artificial Intelligence and Robotics | Operations Research, Systems Engineering and Industrial Engineering

Research Areas

Intelligent Systems and Optimization

Publication

Proceedings of the Twenty-Second International FLAIRS Conference: 19-21 May 2009, Sanibel Island, Florida

First Page

547

Last Page

552

ISBN

9781577354192

Publisher

AAAI Press

City or Country

Menlo Park, CA

Citation

KUMAR, Akshat and ZILBERSTEIN, Shlomo. Dynamic Programming Approximations for Partially Observable Stochastic Games. (2009). Proceedings of the Twenty-Second International FLAIRS Conference: 19-21 May 2009, Sanibel Island, Florida. 547-552.
Available at: https://ink.library.smu.edu.sg/sis_research/2214

Copyright Owner and License

IFAAMAS

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.