Calibration models for determining moisture and fat content of processed cheese using near-infrared spectrometry

The determination of moisture and fat in processed cheese is a common and r egular requirement in the manufacture of this foodstuff; and near-infrared spectrometry in the short-wavelength region (700-1200 mn) can provide the b asis for a suitable on-line and off-line quantitative analytical methodolog y if used with a suitable calibration model. In this study, using data from a 12-filter spectrometer, several calibration models including ordinary le ast squares, multiple linear regression, principal component regression and partial least squares regression have been examined and evaluated for effi cacy in determining moisture and fat content directly and simultaneously in grated cheese samples. Results indicate that orthogonal models using selec ted wavelength data offer superior predictive performance. (C) 1999 Society of Chemical Industry.