Chemometric Methods for the Classification of White Wines

Abstract

Wines are products whose cost depends on several quality factors, required by the customer: geographical origin, variety of grape, oak aging, etc. The Controlled Denomination of Origin (CDO) of wines is frequently desired due to several properties, which depend on characteristics of the different places of origin around the world, including weather, grape variety, crop, temperature variation, winery practices, etc. The control of the properties mentioned above, is usually difficult by traditional methods, since it is necessary to determine several specific variables such as trace elements, organic acids, phenolic compounds, etc., which require expensive equipment, expert operators, long-time analysis and pretreatment of samples, among other undesired aspects. However, the introduction of chemometric tools has simplified the interpretation and analysis of data, allowing the use of a great number of variables (data matrix), filtering only the most important information and leaving out the noisy data. These tools represent a fundamental advantage since they allow the implementation of several spectroscopic methods, which were not useful for such complex analysis: UV-Vis, infrared spectroscopy, nuclear magnetic resonance, and fluorescence methods, which could be used thanks to their multivariable advantage. Based on the benefits of chemometric methods in comparison to traditional methods, this chapter will involve the implementation of the most recent tools used for wine analysis around the world. The determinations include Controlled Denomination of Origin (CDO), geographical and/or botanical origin, and other important quality properties of wines. Chemometric tools include artificial neural network analysis (ANN), Principal Component Analysis (PCA), Cluster Analysis (CA), Linear Discriminant Analysis (LDA), Support Vector Machine (SVM), Partial Least Square Discriminant Analysis (PLS-DA), Soft Independent Modeling of Class Analogy (SIMCA), among others. This chapter has been organized in topics based on the application of methods for quality analysis in white wines, according to botanical and geographical classification, as well as other quality analysis, describing for each one, all the chemometric and analytical methods available for the determination of quality properties, for an easy reading and understanding.