Reinforcement Learning (SS 20)

Lecturer: Jim Mainprice

TA: Philipp Kratzer

All material of this course is available via ilias: Course on ilias


  • Next lectures: July 21 (Q&A via Webex), please prepare some questions
  • The Recap lecture: July 21 will be on webex and streamed/saved on youtube, link was sent via e-mail
  • Next tutorials: July 22 (Inverse RL)

Live Lecture

The lecture will be held using a LIVE STREAM online on the normal lecture slots on the following youtube channel:
There will be a possibility for questions through comments on the live stream. For later access, the videos from the live stream will be uploaded to ilias. The tutorials will be held using Webex, the link and the password is available in ilias.


  • Lectures on Tuesdays 15:45
  • Tutorials on Wednesdays 14:00


Reinforcement Learning is one of the most active research areas in artificial intelligence. It aims to find an optimal policy to achieve a goal by interacting with a complex, uncertain environment – in absence of explicit teachers. Recent success of Reinforcement Learning include mastering the game of GO or learning to play Atari games from raw pixel input. Moreover, there exist various applications in other fields, such as finance, robotics, …


As a prerequisite, students should have knowledge in math (linear algebra, multivariable calculus, statistics, probability theory) and programming (typical programming structures, basic algorithms, object-oriented programming). Coding experience in python is helpful for the exercises. If you are new to python with numpy, we recommend the quickstart.

Previous versions: