Premium
Grey component analysis
Author(s) -
Westerhuis Johan A.,
Derks Eduard P. P. A.,
Hoefsloot Huub C. J.,
Smilde Age K.
Publication year - 2007
Publication title -
journal of chemometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.47
H-Index - 92
eISSN - 1099-128X
pISSN - 0886-9383
DOI - 10.1002/cem.1072
Subject(s) - principal component analysis , computer science , component (thermodynamics) , data mining , interpretation (philosophy) , artificial intelligence , pattern recognition (psychology) , mode (computer interface) , physics , thermodynamics , programming language , operating system
The interpretation of principal component analysis (PCA) models of complex biological or chemical data can be cumbersome because in PCA the decomposition is performed without any knowledge of the system at hand. Prior information of the system is not used to improve the interpretation. In this paper we introduce Grey Component Analysis (GCA) as a new explorative data analysis method that uses the available prior information. GCA uses a soft penalty approach to gently push the decomposition into the direction of the prior information. The grey components are therefore partly data driven and partly driven by the prior information. GCA works in a confirmatory mode to analyze the validity of the prior information and in an exploratory mode in which new phenomena can be studied in detail. To show the wide applicability of GCA, applications within spectroscopy and gene expression are presented. Many diagnostic properties of GCA are introduced and examples of erroneous parts within the prior information are indicated. Copyright © 2007 John Wiley & Sons, Ltd.