Recent Progress in Reinforcement Learning

Not offered this year
Offered occasionally, last taught:

Spring 2021

Reinforcement learning is a framework for studying machines that interact with a sequential environment to achieve a goal. In the past decade, the RL framework has gained a lot of attention owing to its intriguing success in solving problems in complicated domains such as games, robotics, and dialog systems. We observe continual growth in the number of RL papers published in major machine-learning conferences. This growth calls for a careful investigation of the recent progress in the field. By reading selections of the current RL literature, this graduate-level course examines some of the latest theoretical and empirical progress in the field.