Discrimination of adulterants in UHT milk samples by NIRS coupled with supervision discrimination techniques

Abstract

A methodology was developed for distinguishing different ultra-high temperature (UHT) milk adulterants (water, urea, and formaldehyde) at various levels using NIR spectroscopy (NIRS) coupled with supervision discrimination techniques (SIMCA, SVM-DA, and PLS-DA). The figures of merit in discrimination were established based on the results given by the best preprocessing for each technique. Discrimination of formaldehyde was 100% independent of the discrimination technique used. For water discrimination SVM-DA furnished better results as compared to PLS-DA and SIMCA. The same occurred with urea discrimination as well.