Classification of Used Car Prices Using the Naive Bayes Method

Abstract

This research uses the Naive Bayes algorithm to predict used car purchasing decisions based on attributes such as brand, year of production, mileage, engine condition, completeness of features, and maintenance history. By applying the Gaussian Naive Bayes approach to handling continuous data, this research aims to develop a reliable prediction model while identifying the attributes that most influence purchasing decisions. The test results show that the prediction model achieved a correct accuracy level of 80%, and an incorrect accuracy of 20%, which indicates the ability of the Naive Bayes algorithm to handle data classification. This research provides insights that can support industry players in designing more effective sales strategies based on accurate data analysis.