Influence-Based Abstraction for Multiagent Systems
Name
Kaelbling_Influence-based.pdf
Size
159.75 KB
Format
Adobe PDF
Checksum (MD5)
ede3055dc81bad0e203cdee8f3367eff
Author(s)
Oliehoek, Frans A.
Witwicki, Stefan J.
Kaelbling, Leslie P.
Date Issued
July 2012
Journal
Proceedings of the 26th AAAI Conference on Artificial Intelligence
Publisher
Association for the Advancement of Artifical Intelligence
Citation
Oliehoek, Frans A, Stefan J. Witwicki, and Leslie P. Kaelbling. "Influence-Based Abstraction for Multiagent Systems." Proceedings of the 26th AAAI Conference on Artificial Intelligence, July 22-26, 2012, Toronto, Ontario, Canada.
Version
Author's final manuscript
Abstract
This paper presents a theoretical advance by which factored POSGs can be decomposed into local models. We formalize the interface between such local models as the influence agents can exert on one another; and we prove that this interface is sufficient for decoupling them. The resulting influence-based abstraction substantially generalizes previous work on exploiting weakly-coupled agent interaction structures. Therein lie several important contributions. First, our general formulation sheds new light on the theoretical relationships among previous approaches, and promotes future empirical comparisons that could come by extending them beyond the more specific problem contexts for which they were developed. More importantly, the influence-based approaches that we generalize have shown promising improvements in the scalability of planning for more restrictive models. Thus, our theoretical result here serves as the foundation for practical algorithms that we anticipate will bring similar improvements to more general planning contexts, and also into other domains such as approximate planning, decision-making in adversarial domains, and online learning.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
http://www.aaai.org/Conferences/AAAI/2012/aaai12program.pdf
