Automated wine quality assessment using metal-oxide semiconductor sensor arrays and machine learning methods

Wine quality analysis is a vital component of the wine industry, ensuring that the product meets safety and sensory standards. Traditional assessment methods often rely on subjective human judgment and can be inconsistent and time-consuming. This study explores the use of metal-oxide semiconductor (MOS) sensor arrays for an automated and objective analysis of wine quality, focusing on distinguishing fresh wine from spoiled samples. The aim was to identify the presence of acetic acid in wines to prevent spoilage and assess their overall quality. A total of 36 fresh wine samples and 32 spoiled wine samples were continuously monitored for five minutes using the sensor array. The data collected from the sensors underwent noise reduction using discrete wavelet transform, which effectively filtered out irrelevant fluctuations and enhanced signal clarity. For dimensionality reduction, Kernel Principal Component Analysis (KPCA) was applied, reducing the data's complexity while preserving essential information. The reduced data set was then classified using three machine learning algorithms: Support Vector Machine (SVM), k-Nearest Neighbors (KNN), and Decision Tree (DT). The classifiers' performances were evaluated using a 10-fold cross-validation technique to ensure robustness and generalizability. Among the tested models, the SVM classifier exhibited the best performance, achieving an average accuracy of 92.63% in distinguishing between fresh and spoiled wines. The results indicate that the combination of MOS sensor arrays, advanced signal processing, and machine learning algorithms can provide an effective and efficient method of wine quality assessment. The reliability of the results may be affected by the limited number of samples used in this study. This e-nose-based approach offers a reliable alternative to traditional sensory evaluation methods, providing consistent, objective, and rapid analysis, thereby improving quality control processes in the wine industry. The findings suggest that such technological integration could enhance product quality assurance and consumer satisfaction.