MF-OML: Online Mean-Field Reinforcement Learning with Occupation Measures for Large Population Games
This addresses the problem of scalable multi-agent reinforcement learning for large population games, offering the first fully polynomial algorithm for Nash equilibria beyond zero-sum and potential games, which is incremental in extending mean-field methods to broader game classes.
The paper tackles the challenge of computing Nash equilibria in large population sequential symmetric games by proposing MF-OML, an online mean-field reinforcement learning algorithm that achieves approximate Nash equilibria with vanishing gaps as the number of players increases, providing regret bounds of $ ilde{O}(M^{3/4}+N^{-1/2}M)$ and $ ilde{O}(M^{11/12}+N^{-1/6}M)$ under different monotonicity conditions.
Reinforcement learning for multi-agent games has attracted lots of attention recently. However, given the challenge of solving Nash equilibria for large population games, existing works with guaranteed polynomial complexities either focus on variants of zero-sum and potential games, or aim at solving (coarse) correlated equilibria, or require access to simulators, or rely on certain assumptions that are hard to verify. This work proposes MF-OML (Mean-Field Occupation-Measure Learning), an online mean-field reinforcement learning algorithm for computing approximate Nash equilibria of large population sequential symmetric games. MF-OML is the first fully polynomial multi-agent reinforcement learning algorithm for provably solving Nash equilibria (up to mean-field approximation gaps that vanish as the number of players $N$ goes to infinity) beyond variants of zero-sum and potential games. When evaluated by the cumulative deviation from Nash equilibria, the algorithm is shown to achieve a high probability regret bound of $\tilde{O}(M^{3/4}+N^{-1/2}M)$ for games with the strong Lasry-Lions monotonicity condition, and a regret bound of $\tilde{O}(M^{11/12}+N^{- 1/6}M)$ for games with only the Lasry-Lions monotonicity condition, where $M$ is the total number of episodes and $N$ is the number of agents of the game. As a byproduct, we also obtain the first tractable globally convergent computational algorithm for computing approximate Nash equilibria of monotone mean-field games.