High-Dimensional Statistics

Lecturer: Michael Vogt

Excercise: Maximilian Rücker

General informationen:

Lecture:	2 SWS weekly: Wednesday, 10-12 in Heho 22, Lecture Room E.04
Excercise:	2 SWS biweekly: Thursday, 16-18 in Heho 18, Lecture Room 220
Course summary:	The lecture gives an introduction to high-dimensional statistics. Many estimation problems in modern statistics are high-dimensional, that is, the number of parameters to be estimated is much higher than the number of observations. A leading example is the high-dimensional linear regression model, where the number of regressors (and thus the number of parameters) is potentially much larger than the number of data points. There are various statistical methods to estimate the unknown parameters in high-dimensional models such as the lasso and boosting. In the lecture, we derive basic theory for estimation techniques from high-dimensional statistics. We in particular concentrate on theory for the lasso.