Special Lecture (Spezialvorlesung) (E)

Reinforcement Learning (Optimierendes Lernen)
Dr. Joschka Bödecker

• Announcements:
  • first lecture: 18.10.2016
  • Exercises alternating with lectures as announced

• Lectures / Exercises
  • Tuesday, 16:00 - 18:00, bldg. 101 - HS 00 036 SCHICK - SAAL
  • Wednesday, 16:00 - 18:00, bldg. 101 - HS 00 036 SCHICK - SAAL
  • extra lectures will be held on 11/24, 12/1, 12/8 (bldg. 101 - HS 00 036)

• Exam
  • Bachelor: oral exams, March 21st, 2017, from 9:30am (30min slots) in room 00010 of bldg. 079.
  • Master: written exam, March 20th, 2017, 10:00am, Kinohörsaal in bldg. 082

• Credit points:
  • 6 ECTS

• Language:
  • English

The lecture deals with methods of Reinforcement Learning that constitute an important class of machine learning algorithms. Starting with the formalization of problems as Markov decision processes, a variety of Reinforcement Learning methods are introduced and discussed in-depth. The connection to practice-oriented problems is established throughout the lecture based on many examples.

• 11/16, 11/29, 12/7, 12/20, 1/17, 1/24, 2/7

Materials such as slides, assignments, solutions, and recordings are available via ILIAS: RL Course

• The first part of the course is based on the book Neuro-Dynamic Programming by Dimitri P. Bertsekas and John Tsitsiklis (especially chapters 2 and 5).
• Another very accessible book is Reinforcement Learning: An Introduction (link points to the draft of the 2nd edition, to be published in 2017) by Richard S. Sutton and Andrew G. Barto
• A short and concise RL book also containing a lot of material from the course is Algorithms for Reinforcement Learning by Csaba Szepesvári