Application of Pattern Recognition Techniques in the Development of Electronic Tongues
The human tongue, by virtue of the biological receptors present on it, is capable of recognizing five basic tastes: sweetness, saltiness, bitterness, sour, and umami. Likewise, an electronic tongue is a device with synthetic receptors that can distinguish different samples into patterns as a result of the different responses or information extracted from each sample. To perform pattern recognition using an electronic tongue, mathematical tools are required. Hence, chemometric analysis, particularly principal component analysis (PCA), is used for this purpose. PCA is used to represent all data obtained from an electronic tongue using a smaller number of new variables than those in the original data, and representing this new information in 2D or 3D plots. This chapter describes the concept of an electronic tongue, the manner in which PCA is employed to mimic the pattern recognition that occurs in the brain when using the human tongue as a biological receptor, and some applications of such electronic devices for beverage analysis.