Enhancing the Red Wine Quality Classification Using Ensemble Voting Classifiers

Abstract

This study introduces an ensemble voting classifier for red wine quality classification using machine learning algorithms. Wine quality assessment, traditionally reliant on subjective expert evaluations, is addressed through data-driven methodologies. The dataset comprises physicochemical attributes and quality ratings of red wines. Results reveal individual models with accuracy ranging from 0.816 to 0.873, while the ensemble approach significantly enhances accuracy. The combination of Random Forest and XGBoost achieves an accuracy of 0.885, demonstrating its potential in red wine quality assessment. In conclusion, this study showcases the potential of machine learning in enhancing the classification of red wine quality, offering a more objective and precise alternative to traditional sensory evaluation. The ensemble voting classifier, especially when combining Random Forest and XGBoost, provides a robust solution for this task, improving the accuracy of wine quality assessments.