The topic of the seminar is high-dimensional statistics. Roughly speaking, the field of high-dimensional statistics is concerned with estimation problems where the number of parameters to be estimated is much larger than the number of observations. In general, it is hopeless to construct reasonable estimators of the unknown parameters in such models. However, if we impose additional structural constraints on the model, in particular, so-called sparsity constraints, it is possible to do so. In the seminar, we will study statistical methods to deal with (sparse) high-dimensional estimation problems such as the lasso, boosting, random forests and neural networks. The seminar is directed at both Bachelor and Master students. In the winter semester, I also offer a lecture on high-dimensional statistics (Fortgeschrittene Methoden der Mathematischen Biometrie B) which is complementary to the seminar.