Coach-player Multi-agent Reinforcement Learning For Dynamic Team Composition

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

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyliu2021coach

Related papers

• Enhancing Multi-agent Coordination Through Common Operating Picture Integration (2023)0.00
• Emergent Coordination Through Competition (2019)0.00
• Attention-guided Contrastive Role Representations For Multi-agent Reinforcement Learning (2023)3.64
• Collaboration Of AI Agents Via Cooperative Multi-agent Deep Reinforcement Learning (2019)0.00
• Promoting Coordination Through Policy Regularization In Multi-agent Deep Reinforcement Learning (2019)0.00
• Multi-agent Actor-critic For Mixed Cooperative-competitive Environments (2017)0.00
• Hierarchical Cooperative Multi-agent Reinforcement Learning With Skill Discovery (2019)5.24
• Counterfactual Multi-agent Policy Gradients (2017)0.00