<![CDATA[Inventory management at the Genta Sarana Teknik Online Store still relies on simple record-keeping and experience, resulting in frequent instances of sudden stockouts or inventory buildup. This study aims to optimize inventory control by clustering SKUs based on inventory turnover rates using the K-Means algorithm. The research data was derived from SKU master data and inventory transaction history managed in a Java-based desktop application. Each SKU was represented by three numerical features—average stock inflow, average stock outflow, and remaining stock—which were then normalized and processed with K=3 to generate high-, medium-, and low-priority clusters. The final results show that 198 active SKUs were successfully clustered into Cluster 0 (88 SKUs), Cluster 1 (71 SKUs), and Cluster 2 (39 SKUs) with final centroids of 0.102609, 0.284091, and 0.139312; 0.143482, 0.605634, and 0.048502; and 0.473412, 0.700855, and 0.049925. The app provides inventory management features, alternative purchase links for Shopee/Tokopedia, and the clustering process, as well as low-stock alerts and K-Means results. The application of K-Means helps determine restocking priorities in a more objective and structured manner, thereby improving inventory management efficiency]]>
Copyrights © 2026
