Instructor: Rex Liu
This tutorial will provide a crash course on some of the basic methods in reinforcement learning. No prior knowledge beyond Python will be assumed. Emphasis will be on methods rather than proofs. We shall begin with Markov decision processes, the framework upon which all RL is formulated, followed by the central equation that RL essentially attempts to optimise, the Bellman equation. We shall then discuss how all learning methods attempt to optimise this equation, namely through policy evaluation, policy improvement, and value iteration. Finally, we shall cover one of the most important families of RL algorithms, TD-learning, and provide some hands-on exercises to play with these algorithms.
Carney Innovation Space, 4th Floor
164 Angell Street
Pizzas and sodas will be served. Sponsored by the Data Science Initiative and organized by the Center for Computation and Visualization.