Statistics and Its Interface

Volume 6 (2013)

Number 2

Selection consistency of EBIC for GLIM with non-canonical links and diverging number of parameters

Pages: 275 – 284

DOI: https://dx.doi.org/10.4310/SII.2013.v6.n2.a10

Authors

Zehua Chen (Department of Statistics and Applied Probability, National University of Singapore)

Shan Luo (Department of Statistics and Applied Probability, National University of Singapore)

Abstract

In this article, we investigate the properties of the EBIC in variable selection for generalized linear models with noncanonical links and a diverging number of parameters in ultra-high dimensional feature space. The selection consistency of the EBIC in this situation is established under moderate conditions. The finite sample performance of the EBIC coupled with a forward selection procedure is demonstrated through simulation studies and a real data analysis.

Keywords

extended BIC, generalized linear model, non-canonical links, selection consistency, variable selection

2010 Mathematics Subject Classification

Primary 62Fxx. Secondary 62J12.

Full Text (PDF format)

Published 10 May 2013