Ultraviolet-visible spectroscopy and pattern recognition methods for differentiation and classification of wines

The feasibility of developing a qualitative model based on cheap and simple instrumentation to differentiate and classify wines from the apellation d'origine "La Mancha" (Spain) has been studied. The criteria for discrimination were origin, grape variety and ageing process. Ultraviolet-visible spectroscopy was used for the development of a inexpensive and simple screening approach. Once spectra were collected, a data exploratory analysis was carried out in order to both show trends hidden in the data matrix from the sample spectra and study the characteristics of the models thus developed. Principal components analysis and soft independent modelling of class analogy were used for the exploratory analysis and the development of classification models, respectively. The ultraviolet region has for the first time been used for the discrimination of types of wines. This region is of great importance for differentiation of wines according to both origin within a same apellation d'origine and ageing process. The sum of false positives and false negatives - the criteria used for evaluating errors - were not higher than 25%, and the classification of wines according to the origin zone yielded the best values (10%). Sample preparation was not necessary. (c) 2005 Elsevier Ltd. All rights reserved.