ARTIFICIAL NEURAL NETWORK BASED INTELLIGENT TEA TASTER-REVIEW

Анотація

. Tea is the most favorable beverage in the world after the pure water. Professional tea tasters categorize the
quality of the tea in subjective manner by assessing the several parameters. The flavor, aroma and color of tea are the
most 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 which
affect the quality of the tea. Therefore, it is possible to identify a correlation between flavanols composition of tea and
professional tea taster’s valuation. The main purpose of this article is identifying above correlation and according to that
correlation design and implements “Artificial Neural Network (ANN) based Intelligent Tea Taster” to automate manual
tea tasting process. This review is focused on training an artificial neural network according to identified correlation
between flavanols composition and tea taster’s valuation and based on that trained artificial neural network After going
through successful training iterations and evaluations, a computer-based solution can be designed and implemented to
define the quality of tea according to its flavanols compound. The results show good correlation of estimated values of
theaflavins and thearubigins with the actual concentrations obtained by the system when we tested at laboratory. The
review 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.