<![CDATA[Increasing business competition requires companies to understand their product sales patterns in order to make appropriate decisions regarding inventory management and marketing strategies. One approach is to classify products based on their sales performance into best-selling and non-best-selling categories. This study aims to apply the C4.5 algorithm to classify products based on sales transaction data. The dataset consists of sales transaction records with attributes such as product category, price, quantity sold, stock, discount, and total sales. The research stages include data collection, data preprocessing, decision tree construction using the C4.5 algorithm, and model evaluation using a confusion matrix. The results indicate that the C4.5 algorithm is capable of generating easy-to-understand classification rules and achieving a good level of accuracy in determining whether products are best-selling or non-best-selling. The resulting model is expected to assist companies in making better decisions regarding inventory procurement, stock management, and the development of more effective promotional strategies.]]>
Copyrights © 2026
