Fourier transfer infrared (FTIR) spectroscopy is a fast and reliable technique for the authentication of adulteration in the extra-virgin olive oil (EVOO) for quality control and market management. To verify the authenticity of EVOO, the feasibility of FTIR spectroscopy coupled with multivariate calibration has been investigated. Three different multivariate calibrations including linear discrimination analysis, back propagation artificial neural network and least-squares support vector machine (LS-SVM) were studied to make models with the acquired spectra and compared by correlation coefficients in the prediction set. The results demonstrated that the LS-SVM model was superior to others based on its optimum discrimination rates of 100 and 92.5% in the training and prediction set respectively. Besides, all misclassified samples were low-level adulterated EVOO ones with concentrations of 2.5%. This work demonstrates that the FTIR spectroscopy technique combined with an appropriately selected multivariate calibration could be promising to detect of adulterations in EVOO.
展开▼
