Partially Observable Mean Field Multi-Agent Reinforcement Learning Based on Graph-Attention.
Min YangGuanjun LiuZiyuan ZhouPublished in: CoRR (2023)
Keyphrases
- partially observable
- multi agent reinforcement learning
- reinforcement learning
- markov decision processes
- decision problems
- state space
- dynamical systems
- infinite horizon
- stochastic games
- multi agent
- learning agents
- learning agent
- markov random field
- random walk
- reward function
- multi agent learning
- function approximation
- multi agent systems
- optimal policy
- planning domains
- belief state
- em algorithm
- policy iteration
- finite state
- belief networks
- temporal difference
- markov decision process
- dynamic programming
- np hard
- lower bound
- optimal solution
- bayesian networks