Abstract:
Reinforcement Learning (RL) is a framework for sequential decision-making under conditions of uncertainty or incomplete information. It is applicable in various domains, such as autonomous navigation through path-planning in unknown environments, and the management of systems like chemical reactors that are imperfectly modeled. The RL framework also addresses problems in which the information is complete but too voluminous to process by traditional means, such as in games like Chess or Blackjack. This talk will cover the foundational concepts of RL, including states, actions, and rewards, and will discuss the computational approach of solving fixed-point problems to extract practical value from this rich theoretical structure.
Bio:
Dr. Mathukumalli Vidyasagar FRS, a distinguished control theorist and Fellow of the Royal Society, is a Distinguished Professor of Electrical Engineering at IIT Hyderabad. His notable career includes tenures as the Cecil & Ida Green Chair of Systems Biology Science at the University of Texas at Dallas, Executive Vice-President at Tata Consultancy Services, where he headed the Advanced Technology Center, and Director of the Centre for Artificial Intelligence and Robotics (CAIR) at DRDO. His extensive experience and contributions to the field make this session particularly valuable.