After successful completion of the course, students are able to:

• explain the theoretical foundations of machine learning;
• prove learning theoretical results and algorithmic properties of machine learning;
• apply learning algorithms correctly;
• compare and analyse learning algorithms; and
• understand, summarise, and present machine learning research papers.

This lecture introduces theoretical foundations and advanced topics in machine learning. We analyse learning algorithms and show provable guarantees, such as (probabilistic) bounds on the predictive performance.

Tentative topics:

• Empirical risk minimisation and regularisation
• Probably approximately correct (PAC) learning
• VC dimension
• Kernel-based learning (SVM)
• Least squares regression
• Deep learning

A mix of introductory online lectures (recorded and/or live), exercises with formative feedback and some live online sessions where the assigments are discussed.

3ects -> 75h
20h (re)viewing lectures and lecture materials
10h (re)viewing background material
10h exercises
20h coursework
15h Final project

The final grade consists of

• regularly submitted written coursework and
• a larger final project, such as a practical coding project, a written report or a (recorded or online presented) talk