CS60077: Reinforcement Learning

Sequential decision-making is one of the major topics in machine learning. From experience, the task is to decide the sequence of actions to perform in an uncertain environment in order to achieve some goals that may not necessarily seem beneficial in near future but are optimal for getting better long term reward. Reinforcement learning (RL) is a paradigm that proposes a formal framework to this problem. The aim of the course will be to familiarize the students with the basic concepts as well as with the state-of-the-art research literature in deep reinforcement learning. After completion the students will be able to (a) structure a reinforcement learning problem, (b) understand and apply basic RL algorithms for simple sequential decision making problems in uncertain conditions. (c) evaluate the performance of the solution (d) interpret state-of-the-art RL research and communicate their results.