Reinforcement Learning

4 credits

Foundations of modern reinforcement learning. Topics may include Markov decision processes, value iteration, policy iteration, approximate dynamic programming, temporal difference learning, Q-learning, policy gradients, and imitation learning. Recommended: comfortable with programming; and working knowledge of linear algebra (MATH 208); vector calculus (MATH 126); probability and statistics (CSE 312 and STAT 390); algorithms (CSE 421); and machine learning (CSE 446 or CSE 546).