Title: Model-Based and Model-Free: A Tale of Two Paradigms Told from Reinforcement Learning and Generative AI
Abstract: This talk discusses the key connections and differenes between the model-based and model-free paradigms from the perspectives of reinforcement learning and generative AI. It is argued that establishing a sufficiently accurate model is both impossible and unneccessary for the ultimate purpose of making optimal decsions, but there is some quantity, one that is an aggregate measure of the model parameters and control actions, that needs to be learned and can indeed be learned efficiently in a data-driven way.
About the speaker: Xunyu Zhou is the Liu Family Professor of Financial Engineering and the Director of Nie Center for Intelligent Asset Management at Columbia University in New York. His current research focuses on developing a foundational theory for continuous-time reinforcement learning and its applications to financial decision making. Previously, he has worked on quantitative behavioral finance, time inconsistency and stochastic control.
He has addressed the 2010 International Congress of Mathematicians, and has been awarded the Wolfson Research Award from The Royal Society (UK), the Outstanding Paper Prize from the Society for Industrial and Applied Mathematics, the Humboldt Distinguished Lecturer and the Alexander von Humboldt Research Fellowship. He is both an IEEE Fellow and a SIAM Fellow. He was awarded Distinguished Faculty Teaching Award at Columbia University in 2023.
Zhou received his PhD in Operations Research and Control Theory from Fudan University in China in 1989. He was the Nomura Professor of Mathematical Finance and the Director of Nomura Center for Mathematical Finance at University of Oxford during 2007-2016 before joining Columbia.