Numerical Extraction of Components from Mixture Spectra by Multivariate Data Analysis

The authors know by experience that, if presented in purely mathematical terms, the multivariate data analysis techniques which will be discussed in this chapter can be difficult to understand for nonmathematicians. Therefore we will explain techniques such as factor and discriminant analysis, as well as subsequent procedures aimed at extracting chemical information, primarily in geometrical terms. After the basic principles have been visualized, these principles will then be rationalized mathematically so that a link can be established with the more mathematically oriented literature. The paragraphs containing the mathematical rationalizations can be skipped, if desired.