Abstract
The coffee drinking experience is greatly impacted by the aroma it possesses. While human assessors have historically been responsible for classifying the aromas of coffee, recent technological advancements have enabled the utilization of electronic noses (E-noses) to facilitate a more standardized and automated approach. In this paper, we present a novel methodology that combines a separability indicator and a support vector machine (SVM) to effectively select features and optimize the classification process. By testing this method on data from two distinct coffee brands, we observed a significant improvement in classification accuracy. This advancement enhances the performance of sensor data analysis and has the potential to enhance the efficiency and objectivity of coffee aroma classification.