Enhancing red wine quality prediction through Machine Learning approaches with Hyperparameters optimization technique

In light of the intricacy of the winemaking process and the wide variety of elements that could affect the taste and quality of the finished product, predicting red wine quality is difficult. ML methods have been widely used in forecasting red wine quality from its chemical characteristics in recent years. This Paper evaluated the comparison of classification and regression methods to predict the quality of red wine and performed the initial data analysis and exploratory data analysis on the data. This study implemented different Classifiers and Regressors that were trained and tested. Contrasted and Comparative analysis of the accuracies of eight models with hyperparameter tuning optimization, including Logistic Regression, Gradient Boosting, Extra Tree, Ada Boost, Random Forest, Support Vector Classifier, and Decision Tree, Knn and measured the classification report with F1, Accuracy, and Recall Scores. For Imbalance data, SMOTE Classifier was used. This study performed the Cross-validation technique, such as Grid search and with the best hyperparameters tuning. The study's findings demonstrated that the Gradient Boosting technique accurately predicted red wine quality. This research shows the promising results of Gradient Boosting for predicting red wine quality and adds important context to the usage of machine learning classifiers for this challenge.