Reinforcement Learning (RL) describes the highly active and diverse field of learning optimal behavior from interactions with a changing environment, e.g. achieving expert level performance in computer games just by playing them. Conceptually, RL is considered to lie between supervised and unsupervised machine learning approaches and makes use of techniques from both fields.

In this course, you will get to know the tools of modern RL that led to the 2013 breakthrough result of learning to play computer games at super-human level from visual input only. Afterwards, recent and significant improvements to the method as well as current results will be discussed by directly working with the relevant research papers.

After the successful completion of this course, the students

• understand the framework and basic methods for reinforcement learning,
• understand core challenges of the field, and
• will be able to contextualize current publication with respect to those challenges.

Merlin Schüler Lecturer	merlin.schueler@ini.rub.de NB 3/35
Dr. Robin Schiewer Lecturer	robin.schiewer@ini.rub.de

Course type

Seminars

Credits

3

Term

Summer Term 2020

We expect a solid level of mathematics as taught in the Applied Computer Science Bachelor‘s. Tools commonly used in machine learning are

• basic probability theory/statistics (expectations, variance, foundational distributions and densities, markov chains)
• linear algebra (matrices, vectors, eigenvalues/eigenvectors)
• calculus (functions, derivatives/gradients, simple integrals)

The course material is in English, the course language will be English or German – depending on the students.