FM3Q: Factorized Multi-Agent MiniMax Q-Learning for Two-Team Zero-Sum Markov Game.
Guangzheng HuYuanheng ZhuHaoran LiDongbin ZhaoPublished in: CoRR (2024)
Keyphrases
- multi agent
- stochastic games
- markov games
- game theory
- heterogeneous agents
- single agent
- team formation
- multiagent reinforcement learning
- cooperative agents
- american football
- cooperative
- games with incomplete information
- multi agent systems
- multi agent reinforcement learning
- nash equilibria
- repeated games
- imperfect information
- reinforcement learning
- markov decision process
- markov decision processes
- agent cooperation
- multi agent simulation
- multiagent systems
- nash equilibrium
- reinforcement learning algorithms
- game tree
- markov chain
- game theoretic
- perfect information
- team members
- boolean games
- robotic soccer
- extensive form games
- game playing
- incomplete information
- markov model
- autonomous agents
- learning agents
- optimal policy
- worst case
- minimax search
- computer games
- soccer game
- robocup rescue
- temporal difference learning
- virtual world
- robot soccer
- state space
- search algorithm
- coalition formation