Penerapan Algoritma C4.5 untuk Memprediksi Status Penjualan Laptop dan PC pada Toko Elektronik

Abstract

This study aims to apply the C4.5 algorithm to predict the sales status of laptops and PCs in electronics stores to support decision-making in stock management and web-based marketing strategies. The problem faced is the unavailability of a web-based prediction system capable of accurately classifying products into best-selling and non-best-selling categories based on certain attributes. The method used is a data mining technique with a classification approach using the C4.5 algorithm implemented in a web-based system. The research stages include collecting sales data, determining attributes consisting of product type, brand, and price, and forming a decision tree based on the calculation of gain values to generate classification rules. The results show that the product type attribute is the dominant factor in determining sales status. Based on the resulting decision tree, the rule obtained is that products with the PC type are categorized as best-selling, while products with certain types of laptops and brands are categorized as not best-selling, laptops with other specific brands are categorized as best-selling, and laptops with price categories B or C are categorized as not best-selling. The web-based system developed is capable of displaying prediction results quickly and easily understood, so it can be used as a basis for decision-making in electronics stores.