Policy Evaluation in Decentralized POMDPs with Belief Sharing

1 code implementation • 8 Feb 2023 • Mert Kayaalp, Fatima Ghadieh, Ali H. Sayed

As a remedy, we propose a fully decentralized belief forming strategy that relies on individual updates and on localized interactions over a communication network.

Multi-agent Reinforcement Learning