Input variable selection for PLS modeling using nearest correlation spectral clustering

Article ID	Journal	Published Year	Pages	File Type
7563526	Chemometrics and Intelligent Laboratory Systems	2012	11 Pages	PDF

Abstract

âº A new variable selection method for PLS modeling was developed. âº The proposed variable selection method is based on the graph theory. âº The proposed method was applied to softsensor design of an industrial process.

Keywords

Variable selection Spectral clustering Regression Modeling Graph theory

Related Topics

Physical Sciences and Engineering Chemistry Analytical Chemistry

Preview

Input variable selection for PLS modeling using nearest correlation spectral clustering

Authors

Koichi Fujiwara, Hiroshi Sawada, Manabu Kano,