Exploration of Wine features using Data Analytics

The quality of wine is plays a very important role for business owners and customers. The wine quality using traditional method is time consuming and not accurate. To standardize the process of assessing the quality of wine the Machine learning algorithms are frequently used to replace human grading. There are various factors that can be used to predict the wine quality, but not all of them are significant for improved prediction. As a result, the focus of this research is on which wine characteristics are critical for achieving a favorable result. The five methods are used to create the classification model and evaluate the important features: decision tree, random forest, Gradient Boosting, XGBoost, and AdaBoost. The comparison of all five models is shown. Finally, it is demonstrated that the random forest and XGBoost perform better than the decision tree, Gradient Boosting, and AdaBoost in terms of prediction results.