Multi-agent reinforcement learning (MARL) is a varied and highly active field of research that has seen an explosive growth of activity since the mid 2010s. In the wake of this growth, it became clear that the field needed a textbook to provide a principled introduction to MARL for upcoming and seasoned researchers, as well as industry practitioners, and a thorough understanding of the strengths and limitations of our current approaches. In this talk, I will give and introduction to two contributions from the Autonomous Agents Research Group (https://agents.inf.ed.ac.uk/) at the University of Edinburgh to fill this gap.
(1) I will outline the new textbook "Multi-Agent Reinforcement Learning: Foundations and Modern Approaches" (www.marl-book.com), which I co-authored alongside Stefano Albrecht (https://agents.inf.ed.ac.uk/stefano-albrecht/) and Filippos Christianos (https://fchristianos.com/) and will be published with MIT Press in later 2024.
(2) I will give a concise introduction to the EPyMARL codebase (https://github.com/uoe-agents/epymarl) that is the result of a benchmark paper (https://agents.inf.ed.ac.uk/blog/epymarl/) on MARL published at NeurIPS 2021 Track on Datasets and Benchmarks and aims to provide a starting point for cooperative MARL algorithms.
Lukas is a final-year PhD student at the University of Edinburgh under the supervision of Stefano Albrecht. His research focuses on improving the efficiency and generalisability of reinforcement learning, especially in the context of multi-agent systems in which multiple agents have to cooperate. During his PhD, Lukas interned at industry research labs of Dematic, Huawei, and Microsoft Research, and co-authored an introductory open-access textbook on multi-agent reinforcement learning (www.marl-book.com) together with Filippos Christianos and Stefano Albrecht which will be published with MIT Press in late 2024. Prior to his PhD, Lukas graduated with a M.Sc. in Informatics from the University of Edinburgh and a B.Sc. in Computer Science from Saarland University. Lukas is currently in the process of finishing his PhD and looking for postdoctoral research opportunities in decision making.
Note: Please register using the Google Form on our website https://go.umd.edu/marl for access to the Google Meet and talk resources.