부산대학교 도서관

Wine is an alcoholic beverage made from different varieties of grapes after fermenting of grapes. There are different styles of wine varieties depends upon types of grapes used (with or without peeling of skins of grapes) and strains of yeast like red wine, white wine, rose wine, orange or amber wine. Wine can be made from others fruits, grains, cashew coconut, honey etc. Quality and taste of wine depends upon making process and aging of wine. Top 3 Wine producer countries in world are Italy, Spain, and France as per 2021 report. Red wine contains 5.5 to 20.5% of alcohol in it. Consumption of Red wine in right amount is good for health but consumption in large amount is bad for health. The quality of red wine may be predicted by using artificial intelligence techniques with the help of different chemicals parameters as attributes. In this paper, red wine datasets were used for training and testing purpose for the classification of red wine into 6 categories. 6 classes of data were converted into two class based on the quality index. Four different approaches of machine learning algorithms were applied to predict the classification of red wine on two class datasets. It was found that out of four algorithms, the decision tree classifier predicted better performance result of red wine quality compared to other machine learning classifiers.