Wine Quality Detection through Machine Learning Algorithms

Machine learning is one of the emerging areas of research. Many algorithms of data mining have already been used on wine quality dataset to analyze the wine attributes such as quality or class. The quality of wine is not only based on the quantity of alcohol but it also depends on various attributes, these attributes changes with time and so the quality of wine also refines. In this report, machine learning techniques are utilized to analyze those attributes. Firstly data pre-processing takes place i.e. making data appropriate for the models that are built for prediction. Defining independent and dependent variables, missing data handling, feature scaling and data splitting is done to improve the data standard. Then, Logistic regression and Random forest classifier are performed individually on data to predict the test data values. Random forest (RF) classifier outperforms logistic regression (LR) with accuracy 84% while LR has 76% accuracy rate.