Chemometrics and Multivariate Analysis
Published Online: 15 DEC 2009
Copyright © 2003 by Wiley-VCH Verlag GmbH & Co. KGaA. All rights reserved.
Encyclopedia of Applied Physics
How to Cite
Brereton, R. G. 2009. Chemometrics and Multivariate Analysis. Encyclopedia of Applied Physics. 1105–1148.
- Published Online: 15 DEC 2009
A historic introduction to chemometrics is presented. Notation and matrix operations are also presented. Multivariate curve resolution is described, including the use of principal components analysis and the determination of the number of components in a series of mixture spectra. Both noniterative and iterative approaches to resolving out the characteristics (spectra and concentration profiles) of components are outlined. Calibration is covered, including univariate calibration, multiple linear regression, principal components regression, partial least squares, and model validation and assessment. Exploratory data analysis discusses principal components analysis as a method for data visualization and cluster analysis. Supervised methods include ways of developing and assessing classification models, and descriptions of class discriminant analysis, discriminant partial least squares, and K-nearest neighbor method. Finally, multivariate statistical process control is introduced.
- principal components analysis;
- partial least squares;
- multivariate curve resolution;
- pattern recognition