Multi-Agent Learning

Monday, July 1, 2019, 3:00pm – 4:45pm
Presented by Karl Tuyls
Located in 145Π42

The tutorial covers topics in multi-agent learning (MAL). We introduce participants to the very basics, assuming elementary knowledge of single-agent reinforcement learning. We revise some game theoretic concepts and then introduce multi-agent learning, which is non-stationary and reflects a moving target problem, considering several paradigms. We continue by introducing the Markov games multi-agent framework, some elementary game-theoretic solution concepts, and links with replicator dynamics from evolutionary game theory. We review some important algorithms from the ‘pre-deep RL’ period, after which we transition to complex systems and delve into deep multi-agent reinforcement learning. We show illustrations in the AlphaGo, AlphaGo Zero, and Capture the Flag domains, wherein agents learn from scratch and and complex multi-agent behaviors emerge. We conclude by giving an overview of recent developments in ranking of complex multi-agent interactions in large-scale systems.