Classification of Croatian wine varieties using multivariate analysis of data obtained by high resolution ICP-MS analysis

摘要

The main outcome of this work is an elaboration of classification models for seventy-five wine samples representing the most widespread brands of Croatian wines. Analytical measurements are performed by mass spectrometry with inductively coupled plasma, allowing the determination of individual elements. Nine wine varieties originated from three Croatian viticulture regions and thirty-one chemical descriptors (element concentrations of contained elements) are used for their characterization. Different wine categories are characterized by chemometrical techniques, mainly by principal component analysis, cluster analysis, linear discriminant analysis, and analysis of variance. Their role is to discover the elements important for distinct characterization of individual wine categories as well as to assess the possibility of wine samples from different countries being recognized from the increased/decreased content of some elements.