Experience-replay Innovative Dynamics
This work addresses convergence issues in MARL for researchers and practitioners, offering a framework to improve algorithm stability in various game settings, though it is incremental as it builds on existing dynamics.
The paper tackles the instability and nonstationarity in multi-agent reinforcement learning (MARL) by developing a novel experience replay-based algorithm that incorporates revision protocols as tunable hyperparameters, enabling it to mirror trajectories from innovative dynamics like BNN or Smith and extend theoretical guarantees beyond replicator dynamics.
Despite its groundbreaking success, multi-agent reinforcement learning (MARL) still suffers from instability and nonstationarity. Replicator dynamics, the most well-known model from evolutionary game theory (EGT), provide a theoretical framework for the convergence of the trajectories to Nash equilibria and, as a result, have been used to ensure formal guarantees for MARL algorithms in stable game settings. However, they exhibit the opposite behavior in other settings, which poses the problem of finding alternatives to ensure convergence. In contrast, innovative dynamics, such as the Brown-von Neumann-Nash (BNN) or Smith, result in periodic trajectories with the potential to approximate Nash equilibria. Yet, no MARL algorithms based on these dynamics have been proposed. In response to this challenge, we develop a novel experience replay-based MARL algorithm that incorporates revision protocols as tunable hyperparameters. We demonstrate, by appropriately adjusting the revision protocols, that the behavior of our algorithm mirrors the trajectories resulting from these dynamics. Importantly, our contribution provides a framework capable of extending the theoretical guarantees of MARL algorithms beyond replicator dynamics. Finally, we corroborate our theoretical findings with empirical results.