CP 314 : Advanced topics in Robot Learning and Control
Instructors: Shishir N. Y. Kolathaya and Shalabh Bhatnagar
Background: This graduate course will explore the new area of interaction between learning and control specifically applied to robotic systems, both from a foundational level together with a view toward application. The course will first build the necessary framework in which to understand robotic systems, including robot kinematics and dynamics, sensing and estimation, machine learning and control. With these fundamentals the course will focus on data driven approaches for control.
Prerequisites: Students must be well versed with basic mathematical concepts like linear algebra, classical analysis and probability theory. Suggested courses are MA 219, MA 221, MA 222 or MA 361.
Credit Hours: This will be a 3:1 credit course.
Syllabus: Robot dynamics and kinematics, nonlinear control and stability, Markov decision processes, reinforcement learning, imitation learning, model free methods, exploration vs. Exploitation.
Murray, Li and Sastry – A Mathematical Introduction to Robot Manipulation, CRC Press, 1994.
Sutton and Barto Reinforcement Learning – An Introduction, MIT Press, 2017
Sergey Levine: Deep Reinforcement Learning – http://rail.eecs.berkeley.edu/deeprlcourse/
Spong, Hutchinson and Vidyasagar – Robot Modeling and Control, Wiley, 2005.