<![CDATA[Determining seller eligibility in warehouse rental plays a crucial role in maintaining operational stability and minimizing financial risks. However, the selection process is often conducted manually based on subjective judgment, leading to inconsistent and less accurate decisions. This study aims to implement and compare Decision Tree and Random Forest algorithms in predicting seller eligibility using historical data. The dataset consists of 300 records with attributes including Chat Performance, Membership Duration, Rating, and Total Sales. The research process involves data preprocessing, classification model development using RapidMiner, performance evaluation through cross-validation, and feature importance analysis. The results indicate that Random Forest outperforms Decision Tree with an accuracy of 83.11%, while Decision Tree achieves 80.87%. Feature analysis reveals that Chat Performance is the most influential attribute in determining seller eligibility. This research provides a data-driven approach to support objective and consistent decision-making in warehouse rental management.]]>
Copyrights © 2026
