Minimax Sample Complexity for Turn-based Stochastic Game
The empirical success of Multi-agent reinforcement learning is encouraging, while few theoretical guarantees have been revealed. In this work, we prove that the plug-in solver approach, probably the most natural reinforcement learning algorithm, achieves minimax sample complexity for turn-based stochastic game (TBSG). Specifically, we plan in an empirical TBSG by utilizing a `simulator' that allows sampling from arbitrary state-action pair. We show that the empirical Nash equilibrium strategy is an approximate Nash equilibrium strategy in the true TBSG and give both problem-dependent and problem-independent bound. We develop absorbing TBSG and reward perturbation techniques to tackle the complex statistical dependence. The key idea is artificially introducing a suboptimality gap in TBSG and then the Nash equilibrium strategy lies in a finite set.
Nov-28-2020
- Country:
- North America > United States
- California (0.04)
- Europe > United Kingdom
- England > Greater London > London (0.04)
- North America > United States
- Genre:
- Research Report (0.50)
- Industry:
- Leisure & Entertainment > Games (1.00)
- Technology: