University of Vermont

UVM Course Directory

Term: All Terms

Subject: Computer Science

Course Number: 253

CS 253 - QR:Reinforcement Learning

Students will program agents that learn to optimize a reward function using Reinforcement Learning; Markov Decision Processes with discrete states, Value Iteration, Policy Iteration, Q-learning and SARSA, methods for value function approximation in complex domains using linear and non-linear methods. Prerequisites: CS 064 or MATH 052; STAT 151 or STAT 143 or STAT 251; CS 110. Pre/Co-requisites: MATH 122 or MATH 124; CS 125.