A multivariate principal component regression analysis of NIR data

✍ Scribed by Jianguo Sun

• Publisher: John Wiley and Sons
• Year: 1996
• Tongue: English
• Weight: 564 KB
• Volume: 10
• Category: Article
• ISSN: 0886-9383
• DOI: 10.1002/(sici)1099-128x(199601)10:1<1::aid-cem397>3.0.co;2-0

✦ Synopsis

The analysis of near-infrared (NIR) data arising from NIR experiments in which there exists more than one response variable of interest is discussed,with focus on the investigation of the relationship of response variables. A multivariate regression procedure based on principal component regression (PCR), one of the most commonly used methods in NIR analysis, is described. The presented method gives a simultaneous analysis of response variables of interest and is referred to as multivariate principal component regression (MPCR). The idea behind MPCR is the same as that behind PCR, but MPCR could serve as a tool to study the relationship of response variables. MPCR also makes use of the correlation information of the response variables and thus could save a great of computational effort if the response variables are highly correlated. To illustrate MPCR, its application to a set of M R data arising from an NIR experiment is briefly discussed.