Processing of NMR, UV, and IR spectrometric data prior to chemometric simulation by independent component and principal component analysis

Abstract

We consider methods for the mathematical preprocessing of signals in the spectrometric analysis of multicomponent mixtures using chemometric algorithms aimed at adjusting the baseline, experimental noise, and random shift of spectral bands. Practical examples of using simple mathematical operations (scaling, centering, derivatization) are given. The effectiveness of algorithms is illustrated by a wide range of spectroscopic signals (electronic absorption, IR, and NMR spectra) combined with chemometric methods of principal component analysis and independent component analysis.