This talk will start with a discussion of the relationships  between LASSO estimation, ridge regression, and attenuation due to 
measurement error as motivation for, and introduction to, a new  generalizable approach to variable selection in parametric and nonparametric
regression and discriminant analysis. The approach transcends the boundaries of parametric/nonparametric models. It will first be presented in the
 familiar context of linear regression where its relationship to the LASSO will be described. The latter part of the talk will focus on implementation
 of the approach to nonparametric modeling where sparse dependence on predictors is desired. Applications to two-category classification problems
 will be presented.

Authors: Len Stefanski, Yichao Wu, Kyle White
          NC State University