Speaker: Yuming Paul Zhang (Auburn University)

 

Title: Discretization error from regularized Reinforcement Learning to continuous-time stochastic control

 

 

Abstract: While reinforcement learning (RL) typically employs discrete-time Markov Decision Processes (MDPs), its connection to continuous-time optimal control remains a significant theoretical challenge. This work bridges this gap by investigating a class of relaxed control problems with uncontrolled diffusion coefficients. We establish explicit convergence rates for optimal feedback controls across discrete, continuous, relaxed, and classical regimes. If time permits, I will also discuss the convergence properties of the policy iteration algorithm within this framework. These findings provide a rigorous theoretical foundation for implementing RL in stochastic, continuous-time environments.

 

 

 

ACM seminar’s website: https://sites.google.com/view/yzhangpaul/applied-and-computational-mathematics-seminars