Quantitative analysis of near infrared spectra by wavelet coefficient regression using a genetic algorithm

In this paper, we present wavelet coefficient regression (WCR) in combinati on with a genetic algorithm (GA) as a method for multicomponent analysis by Near Infrared Spectrometry. The results are compared with other multivaria te calibration methods like Fourier coefficient regression (FCR), principal component regression (PCR) and absorbance value regression at selected wav elengths (AVR). It is shown that in comparison to conventional methods, WCR is quite unique by the fact that it is self-adaptive. This means that the steps of pretreatment, selection of specific wavelength regions and calibra tion are performed automatically in one step. (C) 1999 Elsevier Science B.V . All rights reserved.