Learning Nash Equilibria in Zero-Sum Stochastic Games via Entropy-Regularized Policy Approximation.
Yue GuanQifan ZhangPanagiotis TsiotrasPublished in: IJCAI (2021)
Keyphrases
- stochastic games
- nash equilibria
- average reward
- multiagent reinforcement learning
- markov decision processes
- incomplete information
- game theory
- nash equilibrium
- repeated games
- rl algorithms
- multi agent
- multi agent reinforcement learning
- game theoretic
- infinite horizon
- optimal policy
- learning automata
- robust optimization
- imperfect information
- reinforcement learning
- pure strategy
- genetic algorithm
- single agent
- state space
- learning agent
- reinforcement learning algorithms
- reward function
- learning tasks
- monte carlo