Efficient Model-based Multi-agent Reinforcement Learning via Optimistic Equilibrium Computation

We consider model-based multi-agent reinforcement learning, where the environment transition model is unknown and can only be learned via expensive interactions with the environment. We propose H-MARL (Hallucinated Multi-Agent Reinforcement Learning), a novel sample-efficient algorithm that can efficiently balance exploration, i.e., learning about the environment, and exploitation, i.e., achieve good equilibrium performance in the underlying general-sum Markov game. H-MARL builds high-probability confidence intervals around the unknown transition model and sequentially updates them based on newly observed data. Using these, it constructs an optimistic hallucinated game for the agents for which equilibrium policies are computed at each round. We consider general statistical models (e.g., Gaussian processes, deep ensembles, etc.) and policy classes (e.g., deep neural networks), and theoretically analyze our approach by bounding the agents' dynamic regret. Moreover, we provide a convergence rate to the equilibria of the underlying Markov game. We demonstrate our approach experimentally on an autonomous driving simulation benchmark. H-MARL learns successful equilibrium policies after a few interactions with the environment and can significantly improve the performance compared to non-optimistic exploration methods.

Automated summary

In this study, we investigate model-based multi-agent reinforcement learning, where the environment transition model is unknown and can only be learned through interactions with the environment. We propose H-MARL (Hallucinated Multi-Agent Reinforcement Learning), an efficient algorithm that balances exploration and exploitation in a general-sum Markov game. By constructing high-probability confidence intervals and updating them based on new data, H-MARL creates an optimistic hallucinated game for the agents to compute equilibrium policies. Experimental results on an autonomous driving simulation benchmark demonstrate that H-MARL learns successful equilibrium policies with only a few interactions and significantly improves performance compared to non-optimistic exploration methods.