Speaker: 김연응 교수 (Gachon University) Date: June 24th.(Fri.) 2022, 16:00 Place: 31351A (SKKU, Natural Sciences 1 Bd. 3F) Title: Discrete optimal control - Learning LQR Abstract: Linear Quadratic Regulator (LQR) is one of the most popular structures in optimal control problems. Recently learning stochastic LQR has received a great deal of attention due to its unique feature. In this talk some basic properties of LQR problems as well as results of regret analysis for learning stochastic LQR are discussed. Particularly we will focus on exploring learning algorithms based on Optimism Face of Uncertainty(OFU) and Thompson Sampling, i.e. posterior sampling.