MIS772 Predictive Analytics: Wine Rating Prediction using Data Mining

This assignment, part of the MIS772 Predictive Analytics course, focuses on developing a data mining model to estimate and classify imported wines based on price categories. The project utilizes a sample of 130,000 wine tasting results to create predictive models using k-NN, decision tree, and gradient boost tree algorithms. The models are evaluated using accuracy, classification error, Kappa, and R2 statistics. The k-NN model is identified as the best fit and is further refined. The assignment also explores the application of RapidMiner for model creation, evaluation, and deployment. The analysis involves examining the relationship between wine ratings, prices, and country of origin. The results demonstrate how predictive analytics can be used to predict wine ratings and provide insights for wine importers. The project also includes the benefits of data analysis and mining for large data sets.