The course will cover topics in statistical learning including:

• regularized regression (lasso and elastic net),
• model assessment and selection (cross-validation and bootstrap),
• regression and classification trees,
• ensemble methods (bagging, random forests, boosting),
• deep learning with neural networks.

Students know about advantages and pitfalls of various statistical learning methods as well as their application and tuning and are able to apply selected methods.

For this course participation is obligatory. Students are allowed to miss a maximum of 20% (no matter if excused or not excused).

This course is taught as lectures and tutorials combined with assignments and a project. In combination with the lectures, the assignments and the project help students to consolidate and expand their understanding of the methods discussed in the lectures.

Grading is based on assignments (30%), a project (30%) and a presentation (40%).