Dr. Jürg Schelldorfer

Duties

• Sep 2007 - Aug 2008 member of the statistical consulting group at the Seminar für Statistik
  
• Sep 2008 - Dec 2011 PhD student and assistant at the Seminar für Statistik

Research Group

I was a member of the German-Swiss Research Group FOR916: Statistical Regularization and Qualitative Constraints.

Visiting Researcher

• (2010) TU Berlin, Machine Learning Group, Prof. Dr. Klaus-Robert Müller, Berlin Brain Computer Interface (BBCI)
• (2010) ZEW Mannheim, Dr. Stephan Dlugosz, Labour Markets, Human Resources and Social Policy (collaboration within the Research Group FOR916)
  

Publications

Preprints:

• Schelldorfer, J., Bühlmann,P. (2011). GLMMLasso: An Algorithm for High-Dimensional Generalized Linear Mixed Models Using L1-Penalization. Preprint arXiv:1109.4003v1

Publications:

• Schelldorfer, J., Bühlmann, P., van de Geer, S. (2011). Estimation for High-Dimensional Linear Mixed-Effects Models Using \ell_1$-Penalization. Scandinavian Journal of Statistics, Vol. 38, 197–214. (doi:10.1111/j.1467-9469.2011.00740.x), (arXiv:1002.3784v2)

• Fazli, S., M. Danóczy, Schelldorfer, J., Müller, K.-R. (2011). l_1-penalized linear mixed-effects models for high-dimensional data with application to BCI. NeuroImage. Vol. 56, Issue 4, 2100-2108. (doi:10.1016/j.neuroimage.2011.03.061)

Ph.D. Thesis:

• High-Dimensional Gaussian and Generalized Linear Mixed Models (doi:10.3929/ethz-a-007018335) under the supervision of Peter Bühlmann
  

Posters:

• Variable Screening and Parameter Estimation for High-Dimensional Generalized Linear Mixed Models Using l_1-Penalization (PDF) at the 2011 ISI Young Statisticians Meeting (YSI 2011)
• lmmlasso: Estimation for High-Dimensional Linear Mixed-Effects Models Using $\ell_1$-Penalization (PDF) ; presented at the inspection of the FOR916
• Using Linear Mixed-Effects Models for subject-independent SMR-base BCI classification (PDF) ; presented at the BCI meeting 2010
  

Talks:

• High-Dimensional Gaussian and Generalized Linear Mixed Models (PDF), Ph.D. talk at ETH Zurich.
  
• Variable Screening and Parameter Estimation for High-Dimensional Generalized Linear Mixed Models Using l_1-Penalization (PDF) at the useR! 2011 in Coventry.
  
• An algorithm for high-dimensional generalized linear mixed models using l_1-penalization (PDF) at the For916 workshop in Mannheim.
  
• Estimation for a high-dimensional mixed-effects model using l₁-constraints (PDF) at the FOR916 workshop in Berne.
• Linear mixed effects models for zero-training BCI (PDF) at the Machine Learning Group at the TU Berlin.

Proceedings:

• Fazli, S., M. Danóczy, Schelldorfer, J., Müller, K.-R. (2011). l_1-Penalized Linear Mixed-Effects Models for BCI. Artificial Neural Networks and Machine Learning - ICAAN 2011.
  

Books containing parts of my research:

• Bühlmann, P. and van de Geer, S. (2011). Statistics for High-Dimensional Data: Methods, Theory and Applications. Springer.

R packages

• lmmlasso: This R package fits (Gaussian) linear mixed-effects models for high-dimensional data using a Lasso-type approach for the fixed-effects parameter (Manual). It can be downloaded from CRAN as well as from R-Forge. How to install it from R-Forge (txt)?
• glmmlasso: This R package fits generalized linear mixed models using a Lasso-type approach for the fixed-effects parameter (Manual). It can be downloaded from R-Forge and here.