CPSC 533-V    Learning to Move    January 2020

This course is about learning to control the movement of humans, animals, and robots, with application to character animation, computer vision, robotics, and biological motor control. Much of the course will focus on (deep) reinforcement learning, which has seen many advances over the past 5 years.

Topics include kinematic and dynamic models of motion, basics of physics-based simulation, classical control methods, dynamic programming, and deep reinforcement learning. Background material will be introduced as necessary. Prior experience in computer graphics, robotics, introductory reinforcement learning, and deep learning will be helpful, but is not necessary.

Lectures Tue Thu 11-12:20pm DMP 101        Jan 7 - Apr 7, 2020
Grading Assignments (45%); Readings, Presentations, and Discussion (20%); Project (35%)
Instructor Michiel van de Panne office hrs: Wed 3-4:30pm, or by appt (ICCS x865)
Resources Questions and discussion: CPSC 533V on Piazza     RL resources
Lectures lectures    
Assignments a1: reading     a2: simulation     a3: tabular Q learning     a4: DQN     a5: PG-PPO     paper presentations     project.pdf