Developing an Online Learning approach to Multi-agent Systems: Game Dynamics and Equilibria

Crucial social and economic settings increasingly involve interactions among machine learning algorithms. This project will investigate the behavior of such complex multi-agent systems by integrating game theory and online learning. Game theory is a long-established mathematical framework for examining strategic interactions, while online learning offers a formal framework for analyzing decision-making in machine learning systems. The project will support two PhD students and two postdocs.