Application of intelligent classifier for multi-dimensional identification of food and beverages

Mariyana Sestrimska, Tanya Titova and Veselin Nachev
University of Food Technologies, Department of Electrical Engineering, Electronics and Automation, Plovdiv, Bulgaria

Abstract

Sestrimska, M., Titova, T. & Nachev, V. (2025). Application of intelligent classifier for multi-dimensional identification of food and beverages. Bulg. J. Agric. Sci., 31(6), 1234–1244

This publication presents models of intelligent recognition systems aimed at the non-destructive quality of food and beverages, in particular, fruit yogurts and tea blends with herbal and fruit ingredients. The experimental study is based on obtaining diffuse reflectance spectral curves for the tea samples and measuring basic physicochemical parameters, such as pH, total soluble solids, and color, for the sour milks.
Naive Bayes Classifier, Back-propagation Artificial Neural Network, and Self-Organizing Map models were synthesized, and the highest classification accuracy was achieved with PC-BP-ANN.

Keywords: Naive Bayes classification; Back-propagation Artificial Neural Network; Self-Organizing Map; PCA; yogurt; tea

See the article as a PDF.