<![CDATA[Uncertainty in stock management, such as the risk of overstock and stock-out, is a major challenge for Office Stationery (ATK) distributors in facing fluctuating seasonal demand patterns in the Business-to-School (B2S) segment. This study aims to analyze the comparative attribute selection criteria in the C4.5 algorithm, namely Information Gain and Gain Ratio, in the classification of best-selling ATK products. The dataset used consists of 647 sales transaction data from January to December 2024. The novelty of this study lies in the comparative analysis of the two criteria in a sales dataset with specific seasonal characteristics, which has not been widely discussed in previous studies that generally only focus on the application of a single algorithm. The research methodology follows the Knowledge Discovery in Database (KDD) stages systematically. The results show that Information Gain produces a slightly higher accuracy value, namely 78.98%, while Gain Ratio (77.89%) produces a model with a simpler, more stable, and easier to interpret decision tree structure. The Procurement Type attribute is identified as the most dominant factor in determining the level of product sales. As a main conclusion, this study establishes that Gain Ratio is a more optimal method for strategic business decision making because through Split Information normalization, this method successfully reduces bias towards highly variable attributes and produces a more concise decision tree structure and avoids overfitting compared to Information Gain.]]>
Copyrights © 2026
