APPLICATION OF RANDOM FOREST IN MACHINE LEARNING TO PREDICT WINE QUALITY

Abstract

Currently, machine learning is applied more and more in life. Machines can
also assist humans in choosing the right products. The producer wants to
produce a suitable wine for the consumer and the customer wants a suitable
wine of his choice. More than half, the quality of wine depends not only on a
certain factor, but it depends on many factors. If you rely on manual methods to
predict the quality, it takes a lot of time. Based on that actual need in this study,
we propose to use 3 methods DT (Decision Tree), SVM (Support Vector Machine),
RF (Random Forest) in machine learning to predict wine. The wine data used as
the basis of the assessment has 1599 lines, each with 12 columns. Experimental
results show that RF method gives the best result, based on this result we build a
wine quality prediction website.