Learning from Multiple Independent Advisors in Multi-agent Reinforcement Learning
Subramanian, Sriram Ganapathi, Taylor, Matthew E., Larson, Kate, Crowley, Mark
–arXiv.org Artificial Intelligence
Multi-agent reinforcement learning typically suffers from the problem of sample inefficiency, where learning suitable policies involves the use of many data samples. Learning from external demonstrators is a possible solution that mitigates this problem. However, most prior approaches in this area assume the presence of a single demonstrator. Leveraging multiple knowledge sources (i.e., advisors) with expertise in distinct aspects of the environment could substantially speed up learning in complex environments. This paper considers the problem of simultaneously learning from multiple independent advisors in multi-agent reinforcement learning. The approach leverages a two-level Q-learning architecture, and extends this framework from single-agent to multi-agent settings. We provide principled algorithms that incorporate a set of advisors by both evaluating the advisors at each state and subsequently using the advisors to guide action selection. We also provide theoretical convergence and sample complexity guarantees. Experimentally, we validate our approach in three different test-beds and show that our algorithms give better performances than baselines, can effectively integrate the combined expertise of different advisors, and learn to ignore bad advice.
arXiv.org Artificial Intelligence
Mar-2-2023
- Country:
- Asia
- Japan (0.04)
- Middle East > Jordan (0.04)
- South Korea > Seoul
- Seoul (0.04)
- Taiwan > Taiwan Province
- Taipei (0.04)
- Europe
- Czechia > Prague (0.04)
- France > Grand Est
- Meurthe-et-Moselle > Nancy (0.04)
- Sweden > Stockholm
- Stockholm (0.04)
- United Kingdom > England
- Cambridgeshire > Cambridge (0.04)
- Greater London > London (0.04)
- North America
- Canada
- Alberta > Census Division No. 11
- Edmonton Metropolitan Region > Edmonton (0.04)
- British Columbia > Metro Vancouver Regional District
- Vancouver (0.14)
- Ontario
- Toronto (0.04)
- Waterloo Region > Waterloo (0.14)
- Alberta > Census Division No. 11
- United States
- New York > New York County
- New York City (0.04)
- Washington > King County
- Bellevue (0.04)
- Massachusetts > Hampshire County
- Amherst (0.04)
- California > Los Angeles County
- Long Beach (0.04)
- Louisiana > Orleans Parish
- New Orleans (0.04)
- Hawaii > Honolulu County
- Honolulu (0.04)
- Nevada (0.04)
- Colorado > Denver County
- Denver (0.04)
- New Jersey > Middlesex County
- New Brunswick (0.04)
- New York > New York County
- Canada
- Oceania
- Australia > Victoria
- Melbourne (0.04)
- New Zealand > North Island
- Auckland Region > Auckland (0.04)
- Australia > Victoria
- South America
- Argentina > Pampas
- Buenos Aires F.D. > Buenos Aires (0.04)
- Brazil > São Paulo (0.04)
- Argentina > Pampas
- Asia
- Genre:
- Research Report > New Finding (0.46)
- Industry:
- Government
- Regional Government (0.70)
- Voting & Elections (0.60)
- Leisure & Entertainment > Games (0.67)
- Government
- Technology: