Impedance modeling for classification of flavored green teas

Abstract

This paper proposes an electrical impedance model of flavored green teas. Typically, impedance data of flavored green teas, obtained by electrochemical impedance spectroscopy (EIS), fit into an equivalent circuit that represents the physical and chemical processes taking place in it. The total impedance of each flavor alone is not sufficient, but different values of impedance parameters in the electrical impedance model are responsible for better classification of flavored green teas. Successfully classified data on the basis of their flavors were obtained by different support vector machine (SVM) techniques with encouraging results. The results show that a linear SVM has better classification results. Analogies between the electrochemical processes within the teas and the electrical impedance model are discussed.