Computational Statistics - How to make complex data analysis feasible Sylvia Fr├╝hwirth-Schnatter

March 25, 2009, 4:15 p.m. HF 9901

Computational statistics is the interface between statistics, computer science and numerical analysis. It aims at developing efficient algorithms that allow to implement statistical methods that were impractical before the advent of computers.

The talk presents several recent research projects, carried out at the Department of Applied Statistics at the Johannes Kepler University. It will be demonstrated how computational statistics makes it possible to analyze increasingly large data sets using statistical methods of increasing complexity. These projects include variable selection and model specification search in high dimensional model spaces with applications in marketing, state space modeling of multivariate time series with applications in finance, and unsupervised clustering of high dimensional data sets with applications in labor economics and bioinformatics.