Optimization of Electrode Array in Electronic Tongue for Classification of Black Tea

Electronic tongue (ET) involves an array of sensors for accurate qualitative analysis of different food products. In this work, five such electrodes out of nine in an ET have been optimized for obtaining reasonable classification accuracy. Herein, nine different electrodes have been prepared using polymer-graphite composites of different compositional ratio. Cyclic voltammogram (CV) has been employed using a three electrode system, and the corresponding response profile has been used for the purpose of data analysis. Further, principal component analysis (PCA) has been applied on the dataset for selecting the appropriate features. The classification accuracy rate of the suitable five electrodes has been selected using support vector machine (SVM) classifier technique keeping the threshold value of accuracy to 90 %.