Reinforcement Learning
Saarland University — Winter Semester 2022

Course structure

This 7-credit course will introduce the basic concepts of Reinforcement Learning (RL) and explore its applications to real-world problems. The course will be structured as follows: (a) the months of November'22 and December'22 will involve weekly assignments comprising of reading material, exercises, and implementation; (b) the month of January'23 will be dedicated to a course project and research papers.

The course will be based on a Flipped Classroom approach and will be primarily based on the book Reinforcement Learning: An Introduction by Sutton and Barto. Every week, we will release weekly assignments that students will have to submit in the following week. There will be a tutorial session held every week where the weekly material will be discussed in detail. In addition, we will schedule designated office hours where students can clarify any questions about the assignments and receive feedback on their solutions. A tentative weekly schedule is as follows:
  • [Tentative] Tutorial sessions: every Tuesday 10:15am - 12pm
  • [Tentative] Office hours: every Thursday 1:15pm - 3pm
  • [Tentative] Release of weekly assignment: Tuesday 1pm with a submission deadline in one week.
  • [Tentative] Submission of assignment: following Monday 6pm (one week after the assignment)

Updates

  • Until 10 October 2022: Register your interest in the course using the registration form provided below.

Course registration

Due to limited capacity in the course, we request you to register your interest using the following Registration Form. In the form, you will have to provide a short motivation text explaining why you are interested in taking this course and mention the relevant courses you have taken along with your grades. We will send you an update about the status of your course enrollment. Please refer to the timeline above.

Organizers

If you have any questions regarding the course, please send an email to adishs@mpi-sws.org.









Imprint / Data Protection