<![CDATA[This study aims to classify the sales performance of food and beverage products using the Decision Tree algorithm. The dataset, obtained from local F&B retail transaction records, includes four key attributes: city, product category, price, and discount percentage, with the target label “Laku” (Yes/No). The research process involves data verification, an 80/20 train–test split, model construction using the Gain Ratio criterion, and performance evaluation through accuracy, precision, and recall metrics in RapidMiner. The resulting model achieved an accuracy of 94.58%, with precision values of 96.53% for the “Laku” class and 89.55% for “Tidak Laku.” The decision rules indicate that price and discount are the most influential factors affecting product sales, where moderate prices and discounts above 10–15% increase the likelihood of products being sold. These findings confirm that Decision Tree provides interpretable and actionable insights for optimizing pricing and promotional strategies. Therefore, the model can serve as a decision-support tool for small and medium F&B enterprises to forecast product performance and enhance marketing decisions.]]>
Copyrights © 2025
