Date and time: 26, 27 and 28 May 2025, 14:00-15:00 CEST
Speaker: Prof. Vivek S. Borkar, Emeritus Fellow, Department of Electrical Engineering, Indian Institute of Technology (IIT), Bombay
Title: Stochastic approximation, Markov decision processes and reinforcement learning
When & Where:
Lecture 1: 26 May, 14:00-15:00 CEST (2:00 to 3:00 PM), Location: H1, KTH main campus
Lecture 2: 27 May, 14:00-15:00 CEST (2:00 to 3:00 PM), Location: Salongen, KTH main campus
Lecture 3: 28 May, 14:00-15:00 CEST (2:00 to 3:00 PM), Location: H1, KTH main campus
To set up research meetings with Prof. Borkar, please contact the organizer of the lectures: Aneesh Raghavan, aneesh@kth.se.
Bio: Prof. Vivek S. Borkar obtained his B.Tech. in Electrical Engg. from IIT Bombay (’76), M.S. in Systems and Control Engg. from Case Western Reserve Uni. (’77) and Ph.D. in Electrical Engg. and Computer Science from Uni. of Californioa at Berkeley (’80). After a visiting stint at the Technical Uni. of Twente (”80-81), he held positions at TIFRCentre (’81-89) and Indian Institute of Science (’89-99) in Bengaluru and TIFR (’99-’11) and IIT Bombay (’11-’20) in Mumbai.
He retired from IIT Bombay in ’20 and held an Emeritus position there till Sept.\ ’24. He has held visiting positions at MIT, Uni. of Maryland at College Park, Uni. of California at Berkeley, and Uni. of Illinois at Urbana-Champaign. He is a Fellow of IEEE, AMS and TWAS as well as various science and engineering academies in India. His research interests are in stochastic control and optimization, with applications to communications and machine learning.
Abstract: The three lectures will cover introduction to stochastic algorithms viewing them as discretizations of continuous dynamics, an overview of Markov decision processes, and reinforcement learning algorithms, a theme that rests upon the preceding two. Each will be covered in a separate lecture in that order, though there may be some overlaps.
