The Complexity of Decentralized Control of Markov Decision Processes
Mathematics of Operations Research
Decentralized Markov Decision Processes with Event-Driven Interactions
AAMAS '04 Proceedings of the Third International Joint Conference on Autonomous Agents and Multiagent Systems - Volume 1
The communicative multiagent team decision problem: analyzing teamwork theories and models
Journal of Artificial Intelligence Research
Perseus: randomized point-based value iteration for POMDPs
Journal of Artificial Intelligence Research
Hi-index | 0.00 |
The Multiagent POMDP (MPOMDP) framework provides well-known methods to model and solve fully communicative multiagent problems. However, the size of these models grows exponentially in the number of agents, and agents are required to act in synchrony. We show how these problems can be mitigated through an event-driven, asynchronous formulation of the MPOMDP dynamics. We can prove that the optimal value function in our framework is piecewise linear and convex, allowing us to extend a standard point-based solver to the event-driven setting. We also show how belief states can be updated at run-time in asynchronous domains. Our results show that asynchronous models scale better to larger domains than synchronous analogues, while retaining solution quality.