Wine Quality Classification

Abstract

This study aims to evaluate red wine quality by analyzing its physicochemical properties. By identifying key measures that significantly influence quality, I aim to provide a systematic assessment of red wine's quality attributes.

Background

This study is centered around the evaluation of red wine quality, a critical factor in the wine industry. The assessment of wine quality is traditionally based on either human sensory analysis or the examination of its physicochemical properties. These properties include crucial factors like pH, dissolved salts, sodium content, acidity, and density, which are known to influence the overall quality of the wine. As the market for high-quality wine continues to expand, there is an increasing demand for more efficient and reliable methods to predict wine quality. Human sensory testing, while insightful, tends to be subjective and time-consuming. Therefore, there is a growing interest in utilizing machine learning techniques within wine informatics to offer a more objective and expedient approach. This involves classifying various wine attributes, including quality, through advanced data analysis, providing a new perspective on quality evaluation beyond traditional methods.

Objective

The goal of this research is to employ a binary classification model to distinguish between high and low-quality white wines. This classification will be based on several critical physicochemical properties of the wines. I am utilizing the red wine quality dataset available from the UCI Machine Learning Repository.

Data

I use the red wine quality dataset from the UCI Machine Learning Repository (http://www3.dsi.uminho.pt/pcortez/wine/winequality.zip) for this analysis.

Colab Notebook

PDF