E-tongue based classification and authentication of mineral water samples following principal component analysis and Sammon's nonlinear mapping

摘要

Present paper proposes the development of E-tongue based water sample classification and authentication using principal component analysis (PCA) enhanced by Sammon's nonlinear mapping method. The eigen value based selection criterion of principal component dimensions in PCA technique has been supplanted by Sammon's mapping in the proposed work. This technique used experimental measurements from a pulse voltametry method which is based on electronic tongue (E-tongue) instrumentation system with silver electrode. The E-tongue system, implemented in a laboratory environment for six numbers of different water samples (Aquafina, Bisleri, Kingfisher, Oasis, Dolphin, and McDowell) available in India was the data source for the present study. The proposed method emancipated encouraging discrimination power with their excellent performances for aforesaid categories of water samples.