SSP'05 -- Proceedings

	IEEE/SP 13th workshop on Statistical Signal Processing July, 17-20, 2005 - Bordeaux - France

Information regarding the paper

Title

A Kullback's symmetric divergence criterion with application to linear regression and time series model

Author(s)

Hocine Belkacemi	LSS/SUPELEC
Abed-Krim Seghouane	National ICT Australia

Get the paper in PDF format

To obtain Acrobat Reader (version 5 minimum required) necessary to his read.

Abstract

The Kullback information criterion (KIC) is a recently developped tool for stistical model selection. KIC serves as an asymptotically unbiased estimator of the Kullback symetric divergence, known as J-divergence. A corrected version for KIC denoted by KICc have been also proposed to corect the bias of KIC. This version tends to overfit when the sample size increases.In this paper we propose an alternative to KICc, the KICu criterion which is unbiased estimator of the Kullback's symmetric divergence. It provides better model choice than KICc for moderate to large sample size.