Scaling Up Multiagent Reinforcement Learning for Robotic Systems: Learn an Adaptive Sparse Communication Graph
Author(s)
Sun, Chuangchuang; Shen, Macheng; How, Jonathan P.
DownloadSubmitted version (469.6Kb)
Open Access Policy
Open Access Policy
Creative Commons Attribution-Noncommercial-Share Alike
Terms of use
Metadata
Show full item recordAbstract
© 2020 IEEE. The complexity of multiagent reinforcement learning (MARL) in multiagent systems increases exponentially with respect to the agent number. This scalability issue prevents MARL from being applied in large-scale multiagent systems. However, one critical feature in MARL that is often neglected is that the interactions between agents are quite sparse. Without exploiting this sparsity structure, existing works aggregate information from all of the agents and thus have a high sample complexity. To address this issue, we propose an adaptive sparse attention mechanism by generalizing a sparsity-inducing activation function. Then a sparse communication graph in MARL is learned by graph neural networks based on this new attention mechanism. Through this sparsity structure, the agents can communicate in an effective as well as efficient way via only selectively attending to agents that matter the most and thus the scale of the MARL problem is reduced with little optimality compromised. Comparative results show that our algorithm can learn an interpretable sparse structure and outperforms previous works by a significant margin on applications involving a large-scale multiagent system.
Date issued
2020-10-24Department
Massachusetts Institute of Technology. Laboratory for Information and Decision SystemsJournal
IEEE International Conference on Intelligent Robots and Systems
Publisher
IEEE
Citation
Sun, Chuangchuang, Shen, Macheng and How, Jonathan P. 2020. "Scaling Up Multiagent Reinforcement Learning for Robotic Systems: Learn an Adaptive Sparse Communication Graph." IEEE International Conference on Intelligent Robots and Systems.
Version: Original manuscript