PENERAPAN KLASIFIKASI ALGORITMA C4.5 BERBASIS KRITERIA PENJUALAN DALAM PENENTUAN STRATEGI PEMASARAN PRODUK HANDPHONE

Abstract

The rapid growth of the mobile device industry has intensified competition among smartphone brands in the retail market, making ineffective inventory management and inappropriate sales strategies potential sources of financial loss. This study aims to classify the sales performance of smartphone models based on brand attributes using the C4.5 decision tree algorithm. The training dataset consists of 41 smartphone models from five major brands (Infinix, Oppo, POCO, Vivo, and Xiaomi), with sales performance categorized into three classes: Low-selling, Moderately Selling, and Best-selling. Mathematical calculations indicate that the system's total entropy is 1.504, while the evaluation of the brand attribute produces an Information Gain value of 0.111. The resulting decision tree reveals that the Infinix, POCO, and Xiaomi brands exhibit complete class purity (Entropy = 0.000), consistently corresponding to the Best-selling category. The implementation of this decision tree enables retail management to predict product sales performance at an early stage, thereby optimizing inventory allocation and improving procurement efficiency.