Premium
Structure modelling and discrimination of Catalan white wines
Author(s) -
Larrechi M. S.,
Franques M. R.,
Ferre M.,
Rius F. X.
Publication year - 2005
Publication title -
journal of chemometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.47
H-Index - 92
eISSN - 1099-128X
pISSN - 0886-9383
DOI - 10.1002/cem.1180030518
Subject(s) - vintage , mathematics , malic acid , artificial intelligence , pattern recognition (psychology) , chemistry , computer science , food science , biochemistry , citric acid
Cluster analysis has been applied to characterized the group structures of four sets of Catalan white wines (Conca de Barberà, Camp de Tarragona, Terra Alta and Ribera‐Falset) on the basis of eleven classical oenological parameters and seven micro and trace metallic constitutents considered to be relatively insensitive to cultural practices. In spite of the vintage variation and the lack of a clear varietal differentiation among the wines, each region could be individually characterized. The application of supervised pattern recognition methods has allowed regional assignment of unknown samples with a prediction rate higher than 95%. Several metal ions (such as calcium, strontium, zinc and magnesium) and a few classical parameters (such as ethanol content and the sum of malic and lactic acid contents) have been found to be relevant for a correct classification. ALLOC and KNN classification methods combined with LDA have been proven useful with the present data set, although their performance was not superior to that of LDA and SIMCA.