Welcome to the Workshop on Location Theory and Robust Statistics 2009

 





Aims and Scops


Location of dimensional structures such as lines, circles, or hyperplanes are a well-researched field in location theory and in computational geometry. On the other hand, the design of robust estimators is of great importance in statistics. Since dimensional structures can be used to approximate a set of points (the simplest example being a regression line) there is a strong relation between both fields. Our workshop aims to point out various relations between robust statistics and the location of dimensional objects. This includes models, algorithms, and results.


Workshop topics include

• location of lines and hyperplanes,
• ordered median problems,
• robust estimators,
• trimmed estimators

with special focus on

• least trimmed squares estimators,
• trimmed likelihood estimators,
• least absolute deviations lines,
• least quantile regression lines,
• LASSO methods,
• numerical approaches,
• discretization approaches.

In order to improve the mutual understanding of researchers in location theory and in robust statistics, the primary goal of this workshop is to learn about the relation between these two fields and to gather knowledge about the state of the art in both fields. Therefore we intend to have extensive discussions. The workshop consists of four invited lectures and some individual contributions.




Local Organizing Committee


Anita Schöbel and Martin Schlather with help of Heike Ahrens, Mark Körner and Daniel Scholz.




Contributions


Participants who wish to present a talk should provide a title and an abstract (maximum one page) until September 25th, 2009. Please send you abstracts by email to Anita Schöbel.




Sponsored by


Graduiertenkolleg “Identification in Mathematical Models: Synergy of Stochastic and Numerical Methods” (DFG)