Classification of Alcoholic Beverage Using Electronic Nose and Support Vector Machine

In recent years, machine learning algorithms have seen significant development and implementation within the technological industry as ways to optimize training data and classification. The study aims to make use of an electronic nose system that is designed to classify five types of alcoholic beverages. Beer, gin, red wine, whiskey, and rum by using their distinct gas signatures. The system uses a Raspberry Pi 4 microcontroller, integrated with MQ gas sensors (MQ-2, MQ-3, MQ-4, MQ-7, and MQ-135) to detect ethanol, carbon dioxide. And other distinct compounds that are emitted by the beverages. A Support Vector Machine (SVM) algorithm was implemented to analyze and classify the data gathered from the sensors. By using 100 trials with 20 samples for each beverage, were used in training and evaluating the model. The results yielded an overall classification accuracy of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 0 \%}$</tex> which was verified by a confusion matrix. The findings demonstrate the potential of the system to identify between different alcoholic beverage types with precision which may have a potential application in quality control and alcoholic beverage authentication.