E-nose Based Classification of Drying Method Using Local Tea Leaves Gas Signature
Meo Vincent C. Caya, Drazen Khristofen R. Romero, Paul A. Castro
Abstract
This study describes the classification of the different drying methods of local tea leaves using support vector machine. The benefit of this study is to give information that shows that there is a practical solution in classifying the drying method used in local tea leaves using an electronic nose and support vector machine. The study resulted in a success where the researchers were able to predict the drying method used on tea leaves by training a model on the raspberry pi and then classifying the data. The researchers obtained an accuracy of 75% for Pandan and 72.22% for Banaba by training the support vector machine algorithm with the collected data. The researchers concluded that the electronic nose system created could predict the drying method based on the gas signatures emitted by the tea leaves using a support vector machine.