Wine Quality Analysis Using Machine Learning Algorithms

Abstract

Wines are being produced since thousands of years. But, it is a complex process to determine the relation between the subjective quality of a wine and its chemical composition. Industries use Product Quality Certification to promote their products and become concern for every individual who consumes any product. It is not possible to ensure quality with experts with such a huge demand of product as it will increase the cost. Wine-makers need a permanent solution to optimize the quality of their wine. This paper explores the space to easy out and make the whole process cost-effective and more trustworthy using machine learning. It allows to build a model with user interface which predicts the wine quality by selecting the important parameters of wine which play a significant role in determining the wines quality. Random forest algorithm is used in determining wines’ quality whose correctness would further be escalated using KNN which makes our model dynamic. Output of this proposed model is used to determine the wines’ quality on a scale of Good, Average or Bad. This proposed model can further be applied to several other products which need quality certification. Our prediction model provides ideal solution for the analysis of wine, which makes this whole process more efficient and cheaper with less human interaction.