A Transfer Learning Framework for Deep Multi-Agent Reinforcement Learning
Dear Editor, This letter presents a new transfer learning framework for the deep multi-agent reinforcement learning (DMARL) to reduce the convergence difficulty and training time when applying DMARL to a new scenario [1], [2]. The proposed transfer learning framework includes the design of neural network architecture, curriculum transfer learning (CTL) and strategy distillation. Experimental results demonstrate that our framework enables DMARL models to converge faster while improving the final performance.
Authors
- Lifeng Ma (ORCID: https://orcid.org/0000-0002-1839-6803)
- Jiacun Wang (ORCID: https://orcid.org/0000-0003-4176-3947)
- Xiang Wu (ORCID: https://orcid.org/0000-0002-9732-628X)
- Yi Liu (ORCID: https://orcid.org/0000-0003-1187-2667)
- Yuming Bo (ORCID: https://orcid.org/0000-0002-0228-6749)
Institutions
- Monmouth University (US)
- Nanjing University of Science and Technology (CN)
Publication Details
- Journal
- IEEE/CAA Journal of Automatica Sinica
- Published
- 2024-10-08
- DOI
- https://doi.org/10.1109/jas.2023.124173
- Citations
- 4
- Primary Topic
- Reinforcement Learning in Robotics
- Type
- article
- Field-Weighted Citation Impact
- 1.39
Funders
- National Natural Science Foundation of China
- Postdoctoral Research Foundation of China