In cooperative multi-agent sequential decision making under uncertainty, agents must coordinate to find an optimal joint policy that maximises joint value. Typical algorithms exploit additive structure in the value function, but in the fully-observable multi-agent MDP (MMDP) setting such structure is not present. We propose a new optimal solver for transition-independent MMDPs, in which agents can only affect their own state but their reward depends on joint transitions. We represent these de- pendencies compactly in conditional return graphs (CRGs). Using CRGs the value of a joint policy and the bounds on partially specified joint policies can be efficiently computed. We propose CoRe, a novel branch-and-bound policy search algorithm building on CRGs. CoRe typically requires less runtime than the available alternatives and finds solutions to previously unsolvable problems.
Original languageEnglish
Title of host publicationProceedings of the Thirtieth AAAI Conference on Artificial Intelligence AAAI-16
PublisherAmerican Association for Artificial Intelligence (AAAI)
Pages3174-3180
Number of pages7
StatePublished - 2016
EventThe Thirtieth AAAI Conference on Artificial Intelligence (AAAI-16) - Phoenix, United States

Publication series

NameProceedings of the AAAI
PublisherAssociation for the Advancement of Artificial Intelligence.
ISSN (Print)2159-5399
ISSN (Electronic)2374-3468

Conference

ConferenceThe Thirtieth AAAI Conference on Artificial Intelligence (AAAI-16)
CountryUnited States
CityPhoenix
Period12/02/1617/02/16

    Research areas

  • Markov Decision Process, Transition-independent Multi-agent MDPs, Reward interactions, Conditional Return Graphs

ID: 10330481