Marten Wegkamp

Professor of Mathematics

411 Malott Hall

(607) 255-7244
marten.wegkamp@cornell.edu

Ph.D. (1996) Leiden University

Research Area

nonparametric statistics

My main research effort concentrates on developing new methodology and rigorous mathematical theory in nonparametric statistics. Currently I am interested in dimension reduction techniques in matrix regression and classification models.

Selected Publications

Complexity regularization via localized random penalties (with Gabor Lugosi), Annals of Statistics 32 no. 4 (2004), 1679–1697.

Weak convergence of empirical Copula processes (with Jean-David Fermanian and Dragan Radulovic), Bernoulli 10 no. 5 (2004), 847–860.

Aggregation for Gaussian regression (with Florentina Bunea and Alexandre Tsybakov), Annals of Statistics 35 no. 4 (2007), 1674–1697.

Uniform central limit theorems for pregaussian classes for functions (with Dragan Radulovic), IMS Collections: High Dimensional Probability V: The Luminy Volume 5 (2009), 84–102.

Optimal selection of reduced rank estimators of high-dimensional matrices (with Florentina Bunea and Yiyuan She), Annals of Statistics 39 no. 2 (2011), 1282–1309.

Support vector machines with a reject option (with Ming Yuan), Bernoulli 17 no. 4 (2011), 1368–1385.