Department of Statistics Seminar
North Carolina State University

presents

Runze Li

The Pennsylvania State University


Multiresponse Surface Models with Block Effects

ABSTRACT

We apply the nonconcave penalized likelihood approach to obtain variable selections as well as shrinkage estimators.  This approach relies heavily on the choice of regularization parameter, which controls the model complexity. In this paper, we propose employing the generalized information criterion (GIC), encompassing the commonly used Akaike information criterion (AIC) and Bayesian information criterion (BIC), for selecting the regularization parameter. Our proposal makes a connection between the classical variable selection criteria and the regularization parameter selections for the nonconcave penalized likelihood approaches.  We show that the BIC-type selector enables identification of the true model consistently, and the resulting estimator possesses the oracle property in the terminology of \citeA{FanLi2001}. In contrast, however, the AIC-type selector tends to overfit with positive probability.  We further show that the AIC-type selector is asymptotically loss efficient, while the BIC-type selector is not.  Our simulation results confirm these theoretical findings, and an empirical example is presented.

Friday, April 3, 2009
3:35 pm--4:35 pm
232A Withers Hall

Refreshments will be served outside the room at 3:00 pm.