Can spectroscopy geographically classify Sauvignon Blanc wines from Australia and New Zealand?

Evaluation of ultraviolet, visible, and near infrared spectroscopy for the analysis of wine compounds

“…In the UV region, 202 and 230 nm were the peaks with the highest responses. These are relative to carboxyl groups of organic acids (Shen et al 2010;Cozzolino et al 2011a;Ferrari et al 2011). Development of multivariate calibration models for pure wine compounds.…”

Differentiation of wines with the use of combined data of UV–visible spectra and color characteristics

“…Geographical discrimination of Spanish wines was demonstrated by the use of UV-visible spectroscopy measurements along with support vector machines to select the relevant wavelengths (Acevedo et al, 2007). In some other studies, combined use of UV-visible range with near and mid infrared spectroscopy was shown to be effective in the classification, requiring more instrumentation other than a UV spectroscopy (Cozzolino et al, 2011). The color parameters accompanied with multivariate statistical methods have been employed in the characterization studies.…”

Review of validation and reporting of non-targeted fingerprinting approaches for food authentication

“…Furthermore, distance measures such as the Mahalanobis distance are used by some of the studies as model diagnostics [54,78,79]. In addition, the Coomans' plot, a graphical illustration for a binary discrimination problem using class distances, was applied in three of the investigated studies [67,80,81]. Other diagnostic plots used for illustrating the classification success are the Y-predict plot [58,65], the distance-to-leverage-plot [80] and the class projection into a principal component score plot [54,82].…”