Learning Nash Equilibria in Zero-Sum Markov Games: A Single Time-scale Algorithm Under Weak Reachability.
Reda OuhammaMaryam KamgarpourPublished in: CoRR (2023)
Keyphrases
- learning algorithm
- stochastic games
- nash equilibria
- computational complexity
- cost function
- dynamic programming
- fictitious play
- incomplete information
- learning process
- objective function
- machine learning
- worst case
- path planning
- markov decision processes
- game theory
- convergence rate
- nash equilibrium
- reinforcement learning