Dipl. Stat. Jan Gertheiss

Forschung / Research

• High Dimensional Data
• Structured Predictors
• Feature Selection

Refereed Papers & Book Chapters:

• Gertheiss, J. and Tutz, G. (2009): Feature Selection and Weighting by Nearest Neighbor Ensembles, Chemometrics and Intelligent Laboratory Systems, 99, 30-38;
  preliminary version: LMU Munich, Department of Statistics: Technical Report 33.
• Gertheiss, J. and Tutz, G. (2009): Penalized Regression with Ordinal Predictors, International Statistical Review, 77, 345-365;
  preliminary version: LMU Munich, Department of Statistics: Technical Report 15.
• Gertheiss, J. and Tutz, G. (2009): Supervised Feature Selection in Mass Spectrometry based Proteomic Profiling by Blockwise Boosting, Bioinformatics, 25, 1076-1077.
• Tutz, G. and Gertheiss, J. (2009): Feature Extraction in Signal Regression: A Boosting Technique for Functional Data Regression, Journal of Computational and Graphical Statistics (to appear);
  preliminary version: LMU Munich, Department of Statistics: Technical Report 11.
• Gertheiss, J. and Tutz, G. (2009): Statistische Tests, in: M. Schwaiger and A. Meyer (eds.): Theorien und Methoden der Betriebswirtschaft, Munich: Vahlen, 439-454.
• Gertheiss, J. and Tutz, G. (2009): Variable Scaling and Nearest Neighbor Methods, Journal of Chemometrics, 23, 149-151.

Preprints:

• Gertheiss, J. and Tutz, G. (2010): Regularization and Model Selection with Categorial Effect Modifiers, LMU Munich, Department of Statistics: Technical Report 73.
• Gertheiss, J., Hogger, S., Oberhauser, C. and Tutz, G. (2009): Selection of Ordinally Scaled Independent Variables, LMU Munich, Department of Statistics: Technical Report 62.
• Gertheiss, J. and Tutz, G. (2009): Sparse Modeling of Categorial Explanatory Variables, LMU Munich, Department of Statistics: Technical Report 60.

• Clustering of Categories in Multiple Regression with Categorial Predictors, ERCIM 2009, October 29-31, Limassol, Cyprus.
• Penalized Regression with Ordinal Predictors, Biometrisches Kolloquium 2009, March 17-19, Hannover, Germany.
• Variable Selection and Weighting by Nearest Neighbor Ensembles, Workshop on Nonparametric Inference 2008, June 26-28, Coimbra, Portugal.
• Feature Extraction in Signal Regression: A Boosting Technique for Functional Data Regression, Lifestat 2008, March 10-13, Munich, Germany.

• BlockBoost R-functions for normal outcomes, as described in Tutz and Gertheiss (2009). An example analysis of the Canadian weather data (Ramsay and Silverman, 2005) using blockwise boosting can be found here.
• BlockBoost R-functions for binary outcomes, based on the LogitBoost R-code (R-package boost 1.0-0, Dettling, 2008), with an example analysis of a subset of the Petricoin et al. (2002) prostate cancer data using blockwise boosting; original data can be found here.