Eliminate indeterminacies of independent component analysis for chemometrics

An improved method has been proposed to eliminate the indeterminacies of independent component analysis (ICA) for chemometrics. Following the arrangement of principal components analysis (PCA), the ICA mixing matrix is selected as signal content indexes, and ICA output are sorted and directed. After...

Full description

Saved in:
Bibliographic Details
Published inProgress in natural science Vol. 18; no. 8; pp. 1009 - 1014
Main Authors Yao, Zhixiang, Zhang, Kai, Liu, Huanbin, Su, Hui
Format Journal Article
LanguageEnglish
Published Elsevier Ltd 2008
Department of Biological and Chemical Engineering, Cruangxi University of Technology, Liuzhou 545006, China
State Key Laboratory of Pulp and Paper-making Engineering, South China University of Technology, Guangzhou 510640, China%State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Beijing 102249, China%State Key Laboratory of Pulp and Paper-making Engineering, South China University of Technology, Guangzhou 510640, China%Department of Biological and Chemical Engineering, Cruangxi University of Technology, Liuzhou 545006, China
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:An improved method has been proposed to eliminate the indeterminacies of independent component analysis (ICA) for chemometrics. Following the arrangement of principal components analysis (PCA), the ICA mixing matrix is selected as signal content indexes, and ICA output are sorted and directed. After many times reputations, independent components (ICs) are paired according to the maximum correlation coefficient, and then the mean values of each IC substitutes the original ICs. This indicates that the ICA indeterminacies are eliminated. A simulation example is tested to validate this improvement. Finally, a set of experimental LC MS data is processed without any prior knowledge or specific limitation and the results show that the improved ICA can directly separate the mixed signals in chemometrics, and it is simpler and more reasonable than the simple to use interactive self-modeling mixture analysis (SIMPLISMA).
Bibliography:Independent component analysis; Indeterminacy in ICA; Chemometrics
O6-04
Indeterminacy in ICA
Chemometrics
Independent component analysis
11-3853/N
ISSN:1002-0071
DOI:10.1016/j.pnsc.2008.01.034