Abstract
The practice of using multivariate methods for spectral chemical analysis has increased. The inflation of computers incorporated with spectroscopic instruments represents the main thrust behind the growth. For example, multivariate procedures are typically applied to Fourier transform infrared (FTIR) spectroscopy and ultraviolet-visible (UV-Vis) spectrophotometry. The most common multivariate routines for chemical analysis consist of multiple linear regression (MLR), principal component regression (PCR), and partial least