Black tea quality evaluation using electronic nose: An Artificial Bee Colony approach

The process of evaluating the quality of black tea by human sensory panel known as “Tea Tasters” is very subjective and its accuracy is questionable. As a result various instrumental setups have been investigated in the past for quality determination of black tea. The electronic nose is one such instrument, consisting of sensor arrays, odor detection and data acquisition system. Many clustering algorithms are available that can be used to differentiate between tea with various scores as given by the “Tea Tasters”, the data for which is obtained by an electronic nose. Recently, a heuristic algorithm called the Artificial Bee Colony (ABC) has been devised by Karaboga mainly for solving optimization problems but which has also been applied for data clustering. In this paper, the algorithm devised by Karaboga has been used on the data obtained by an electronic nose to distinguish among the various scores of black tea. Also the results obtained by ABC have been compared with the well known Fuzzy C-Means (FCM) algorithm. The paper is concluded by some observations made by the authors.