The GML (Grundlagen Maschinelles Lernen) are now substitute by the new lecture by Prof. Besserve, see Maschine Learning for Data Science.

If you did the GML lecture before (up to SS24) with Prof Steil, there is still the possiblity to take the exam, it will take place as a written exam on 28.07.2025.

Some remarks:

• the exam will be in written
• it is a repeat exam and will very similar to the previous exams
• you should use the course material from SS 2024
• all relevant material can be found in StudIP
• on preparation: I have discussed a number of techniques for learning in the lecture. Generally speaking focus on the essential things in the "what is the most important: formula/definition/learning rule/etc."
• in StudIP you have the list of questions, which helps you and guides through the slides
• you need to be able to write down the essential equations, learning rules, and definitions.
• to learn this, it is often helpful to compile a kind of cheat sheet, which has the most important information (but which you can not use in the exam, of course)
• do not go from start to end with all detail - better make sure the basics are clear for all the chapters and only afterwards go to the detail.
• make sure you know the definitions and do not confuse e.g. k-means with KNN
• get the basic math/defintions on concept learning right (it is very simple math, but needs to be precise)