Chemometric classification of Basque and French ciders based on their total polyphenol contents and CIELab parameters

Abstract

Total polyphenol contents, estimated by Folin–Ciocalteu method, and CIELab chromatic parameters were determined in Basque and French ciders with the aim of developing a classification system to confirm the authenticity of ciders. A preliminary study of data structure was performed by a multivariate data analysis using chemometric techniques such as cluster analysis and principal component analysis. Supervised pattern recognition methods, such as linear discriminant analysis, K-nearest neighbours (KNN), soft independent modelling of class analogy and multilayer feed-forward artificial neural networks (MLF-ANN), provided classification rules for the two categories based on the experimental data. KNN results for Basque ciders afforded an excellent performance in terms of recognition and prediction abilities (99%), providing a useful tool to detect genuine Basque ciders. Despite KNN and MLF-ANN giving the best results for French ciders, with a success rate of prediction ability around 91%, this would not be acceptable for authentication purposes.