Skip to main content

Approximate Solutions for Factored Dec−POMDPs with Many Agents – Extended Abstract

Frans Oliehoek‚ Shimon Whiteson and Matthijs Spaan

Abstract

Dec-POMDPs are a powerful framework for planning in multiagent systems, but are provably intractable to solve. This paper proposes a factored forward-sweep policy computation method that tackles the stages of the problem one by one, exploiting weakly coupled structure at each of these stages. An empirical evaluation shows that the loss in solution quality due to these approximations is small and that the proposed method achieves unprecedented scalability, solving Dec-POMDPs with hundreds of agents.

Book Title
BNAIC 2013: Proceedings of the Twenty−Fifth Benelux Conference on Artificial Intelligence
Month
November
Year
2013