Using neural networks for wine identification

Abstract

This paper presents a study and research on mathematical models for the classification of wine absorption spectrum characteristics in order to improve some critical aspects of the production in the bottling process. The classifier model described, based on neural networks, performs both recognition and classification of various typologies of wine produced from the wine firm "M. Mastroberardino" of Atripalda.