<![CDATA[The management of consumable inventory expenditure at the Human Resources and Personnel Development Agency (BKPSDM) of Kudus Regency still encounters challenges due to manual and unsystematic data analysis processes. This study aims to classify inventory usage data using the K-Means Clustering method supported by the CRISP-DM framework. The variables used include the quantity and total value of item expenditures. A web-based system was also developed to support the clustering process and provide analysis results more effectively. The K-Means++ algorithm was implemented to obtain better centroid initialization. The findings show three main clusters representing low, medium, and high expenditure levels. The system presents clustering results in tables and charts, making them easier to interpret for decision-making. This research is expected to support inventory planning and improve efficiency in inventory management at BKPSDM Kudus.]]>
Copyrights © 2026
