Have you always wanted to know how to automatically recognize spam e-mails, classify brain data or sort out faulty products based on camera data? The course introduces you to the research field of machine learning (ML), whose methods provide solutions for these and many other problems, in order to learn connections from existing data. Methods of unsupervised and, above all, supervised learning are dealt with. Topics include: regression and classification, dimensionality reduction, random forests, deep learning, kernel methods and much more.

ML methods are largely responsible for the success of various applications in industry and science and have, for example, revolutionized image and speech recognition in recent years. Due to advances in algorithm development and the associated theory, as well as the steadily increasing amounts of data and increasing computing power, it is foreseeable that ML will continue to gain in importance in the future in order to enable data-driven knowledge gain.

Course participants obtain a fundamental understanding of ML algorithms in theory and practice. You will learn which statistical assumptions are made by different algorithms, which pitfalls to avoid when analyzing data, and how great the effort is both for training ML models and for their productive use. In addition, the participants analyze given data, learn to implement ML methods and use Python-based ML libraries in practice. 

The course takes place in English. Basic knowledge of algorithms and data structures as well as basic mathematical knowledge of analysis and linear algebra are required for this course. Knowledge of the Python programming language is helpful.

You can find more information about the required prior knowledge in the Continuing Education Courses section.

The costs including learning materials, tutoring by a research assistant and the exam amount to 2,300 euros.