“An Introduction to the Theory of Reinforcement Learning (and Why It Must Be Zapped!)”
Thursday, Nov. 7 at 1:00 pm
According to Wikipedia, “Reinforcement learning (RL) is an area of machine learning concerned with how software agents ought to take actions in an environment so as to maximize some notion of cumulative reward.” A more meaningful description comes later in the article: “The main difference between the classical dynamic programming methods and reinforcement learning algorithms is that the latter do not assume knowledge of an exact mathematical model (of the system to be controlled).” In reality, RL is an emerging science for the creation of decision and control strategies for complex systems. The ideas have had tremendous success in games (Google’s famous Go and chess algorithms), and potential value in a range of applications, especially in robotics. This talk will provide a gentle introduction to the main ideas, some recent success stories at UF, and a list of some of the challenges we now face. The success stories require an introduction to dynamic programming and stochastic approximation (a foundation of many algorithms in machine learning).
- Poetry from Google: https://www.vox.com/future-perfect/2019/9/20/20872672/ai-learn-play-hide-and-seek
- Science from UF: https://arxiv.org/abs/1707.03770
Dr. Sean Meyn is Professor and Robert C. Pittman Eminent Scholar Chair at the Department of Electrical and Computer Engineering at the University of Florida. His research focuses on Markov processes (with or without control), spectral theory and large deviations; stochastic approximation, and reinforcement learning. He received his Ph.D. in Electrical Engineering from McGill University in 1987.