Settling the Variance of Multi-Agent Policy Gradients

Jakub Grudzien Kuba, Muning Wen, Linghui Meng, Shangding Gu, Haifeng Zhang, David Henry Mguni, Jun Wang, Yaodong Yang

Keywords: Multi-Agent Reinforcement Learning, Policy Gradient, Variance Reduction, Statistics

TL;DR: The paper analyses the variance of multi-agent policy gradient estimates, and provides a technique to reduce it.

Abstract: Policy gradient (PG) methods are popular reinforcement learning (RL) methods where a baseline is often applied to reduce the variance of gradient estimates. In multi-agent RL (MARL), although the PG theorem can be naturally extended, the effectiveness of multi-agent PG (MAPG) methods degrades as the variance of gradient estimates increases rapidly with the number of agents. In this paper, we offer a rigorous analysis of MAPG methods by, firstly, quantifying the contributions of the number of agents and agents' explorations to the variance of MAPG estimators. Based on this analysis, we derive the optimal baseline (OB) that achieves the minimal variance. In comparison to the OB, we measure the excess variance of existing MARL algorithms such as vanilla MAPG and COMA. Considering using deep neural networks, we also propose a surrogate version of OB, which can be seamlessly plugged into any existing PG methods in MARL. On benchmarks of Multi-Agent MuJoCo and StarCraft challenges, our OB technique effectively stabilises training and improves the performance of multi-agent PPO and COMA algorithms by a significant margin. Code is released at \url{https://github.com/morning9393/Optimal-Baseline-for-Multi-agent-Policy-Gradients}.

Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.

Code: https://github.com/morning9393/Optimal-Baseline-for-Multi-agent-Policy-Gradients

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2108.08612/code)