Department of Statistics Seminar
North Carolina State University

presents

Dr. Yi Lin

University of Wisconsin

"Component Selection and Smoothing in Smoothing Spline Analysis of Variance Models"

ABSTRACT

In this talk I introduce a new method for model selection and model fitting in nonparametric regression models, in the framework of smoothing spline analysis of variance. The "COSSO" is a method of regularization with the penalty functional being the sum of component norms, instead of the squared norm employed in the traditional smoothing spline method. The COSSO provides a unified framework for several recent proposals for model selection in linear models and smoothing spline analysis of variance models. Theoretical properties in terms of estimation and model selection are studied. The COSSO is compared with the MARS in simulations and real examples, and is shown to give very competitive performances.

Friday, February 7, 2003

3:35 - 4:35 pm

206 Cox Hall

Refreshments will be served on the second floor of Dabney Hall (left of Room 222) at 3:00 pm.