Coach-Player Multi-Agent Reinforcement Learning for Dynamic Team Composition

TL;DR

This paper introduces COPA, a coach-player multi-agent reinforcement learning framework that effectively manages dynamic team compositions with partial information, achieving strong generalization and communication efficiency.

Contribution

The paper presents a novel coach-player framework with attention, variational regularization, and adaptive communication for dynamic multi-agent team coordination.

Findings

Achieves zero-shot generalization to new team compositions.

Performs comparably or better than full-view baselines.

Maintains high performance with minimal communication (13%).

Abstract

In real-world multi-agent systems, agents with different capabilities may join or leave without altering the team's overarching goals. Coordinating teams with such dynamic composition is challenging: the optimal team strategy varies with the composition. We propose COPA, a coach-player framework to tackle this problem. We assume the coach has a global view of the environment and coordinates the players, who only have partial views, by distributing individual strategies. Specifically, we 1) adopt the attention mechanism for both the coach and the players; 2) propose a variational objective to regularize learning; and 3) design an adaptive communication method to let the coach decide when to communicate with the players. We validate our methods on a resource collection task, a rescue game, and the StarCraft micromanagement tasks. We demonstrate zero-shot generalization to new team compositions. Our method achieves comparable or better performance than the setting where all players have a full view of the environment. Moreover, we see that the performance remains high even when the coach communicates as little as 13% of the time using the adaptive communication strategy.