Module Number

ML-4350
Module Title

Reinforcement Learning
Lecture Type(s)

Lecture, Tutorial
ECTS 6
Work load
- Contact time
- Self study
Workload:
180 h
Class time:
60 h / 4 SWS
Self study:
120 h
Duration 1 Semester
Frequency In the summer semester
Language of instruction English
Type of Exam

Presentation and written report (or exam, will be determined)

Content

The lecture covers the whole range of reinforcement learning topics, from the basic formalism and theory to state-of-the-art algorithms.
The lecture is accompanied by exercises that help to deepen the mathematical understanding, as well as hands-on experience in implementing algorithms.

• Introduction to supervised learning and optimization
• Basics of Reinforcement Learning (RL) and Markov Decision Processes
• Dynamic programming, prediction and control
• Value function approximation
• Policy gradient
• Deep RL, control in continuous state-action domains
• Optimal control and model-based RL
• Advanced topics in RL

Objectives

(1) Students can phrase a problem in the reinforcement learning framework and
can select an appropriate algorithm for solving it.
(2) Students are able to implement a set of deep reinforcement learning algorithms
and analyse their behavior.
(3) Students can explain the challenges in reinforcement learning and assess
and characterize new reinforcement learning methods.

Allocation of credits / grading
Type of Class
Status
SWS
Credits
Type of Exam
Exam duration
Evaluation
Calculation
of Module (%)
Prerequisite for participation There are no specific prerequisites.
Lecturer / Other Martius
Literature

Reinforcement learning by Sutton and Barto http://incompleteideas.net/
book/bookdraft2017nov5.pdf
Pattern Recognition and Machine Learning by C.M. Bishop, Chap. 3 and 5
Deep Learning by Goodfellow, Bengio and Courville https://www.
deeplearningbook.org / Recommended to attend basic Machine learning class before.

Last offered Wintersemester 2021
Planned for Sommersemester 2023
Assigned Study Areas INFO-INFO, INFO-PRAK, INFO-THEO, MEDI-APPL, MEDI-INFO, ML-CS, ML-DIV