Читать книгу Multiblock Data Fusion in Statistics and Machine Learning - Tormod Næs - Страница 32

Multiblock data analysis problems arise in different parts of chemistry. A very active area is analytical chemistry, with two very prominent topics. The first one is multivariate curve resolution where the general idea is to mathematically resolve chemical mixtures in underlying pure chemical components and their concentration profiles (Tauler et al., 1995; de Juan and Tauler, 2006). Many different types of multiblock data analyses are performed in this area with a special emphasis on applying domain-specific constraints. The second application area is calibration where the purpose is to obtain concentrations from instrumental analysis methods. Also in this area multiblock data analysis methods are used (Næs et al., 2013). A spectroscopy example is given in Example 1.3.