Short Course Description
Online Learning and Reinforcement Learning are two subdomains of Artificial Intelligence.
In Online Learning, model training is continuous and in some cases done even when the training examples are generated by an adversary. The goal in Online Learning is to design learning algorithms that make step-by-step predictions and minimize the overall "regret" --- namely, make sure that the prediction error is as close as possible to that of the best in hindsight.
In Reinforcement Learning, on the other hand, the goal is to generate long-term decisions.
In this course, we delve into these two fields and observe how, in recent years, tools from Online Learning have been used to solve a diverse set of problems in Reinforcement Learning and in Control Theory.
Full Syllabus
