Artificial Neural Network Based Intelligent Tea Taster-Review

Abstract

. Tea is the most favorable beverage in the world after the pure water. Professional tea tasters categorize thequality of the tea in subjective manner by assessing the several parameters. The flavor, aroma and color of tea are themost important and considered parameters when professional tea tasters categorize and evaluate tea. The value of abovementioned parameters depends on the chemical composition of the tea. Basically, flavanols are major compounds whichaffect the quality of the tea. Therefore, it is possible to identify a correlation between flavanols composition of tea andprofessional tea taster’s valuation. The main purpose of this article is identifying above correlation and according to thatcorrelation design and implements “Artificial Neural Network (ANN) based Intelligent Tea Taster” to automate manualtea tasting process. This review is focused on training an artificial neural network according to identified correlationbetween flavanols composition and tea taster’s valuation and based on that trained artificial neural network After goingthrough successful training iterations and evaluations, a computer-based solution can be designed and implemented todefine the quality of tea according to its flavanols compound. The results show good correlation of estimated values oftheaflavins and thearubigins with the actual concentrations obtained by the system when we tested at laboratory. Thereview is based on ANN base Intelligent Tea Taster will automate the tea tasting process while improving efficiency,effectiveness and accuracy of the tea tasting process.