Chemometric strategies for authenticating extra virgin olive oils from two geographically adjacent Catalan protected designations of origin

This paper proposes using multivariate methodology to authenticate the geographical origin of monovarietal Arbequina extra virgin olive oils from two geographically adjacent Catalan PDOs (protected designations of origin). Two spectroscopic techniques, fluorescence and FT-Raman were used to obtain multivariate data. The results obtained by principal component analysis (PCA), partial least squares-discriminant analysis (PLS-DA) and low and high-data fusion are discussed. The training and test samples were randomly chosen. To obtain greater representativeness in the results, the data set was independently divided three times. Thus, three different training and test sets were obtained and used to build independent classification models and carry out data fusion strategies. When the two fusion strategies were applied, the performance parameters (sensitivity and specificity) were better than when they were applied individually, which indicates that synergies between FT-Raman and fluorescence expand the information about the samples. It was also observed that the influence on the results of the type of data set obtained by the random division of the data set was minimized.

Automated summary

This study proposes using multivariate methodology combined with spectroscopic techniques to authenticate the geographical origin of monovarietal olive oils, showing that fusion strategies can improve identification performance, and the way the dataset is divided can influence the results.