Detection and quantification of cow milk adulteration using portable near-infrared spectroscopy combined with chemometrics

Milk adulteration is a common phenomenon in many countries, which draws extensive attention from humans due to health hazards that might result some fatal diseases. In this study, portable near-infrared (NIR) spectrometer combined with multivariate analysis was used detect and quantify milk adulteration. Fresh cow samples were collected eight dairy farms Beijing Hebei province of China. Water, urea, starch goat adulterate at 11 different concentrations. The data driven soft independent modeling class analogy (DD-SIMCA) method employed for qualitative analysis. Partial least squares regression (PLSR) applied statistical the obtained NIR spectral data. results showed DD-SIMCA approach achieved satisfactory classification. By PLSR model, standard error prediction (SEP) values 4.35, 0.34, 4.74 5.56 g/L water, milk, respectively. These demonstrated feasibility reliability spectroscopy total contents investigated adulterants milk. Key words: Portable (NIR)-spectroscopy, adulteration, DD-SIMCA, (PLS) regression.