<![CDATA[Human Resources (HR) are a strategic asset in institutional advancement, so employee performance evaluation must be conducted objectively and based on data. This study aims to cluster employee performance data at XYZ University for determining further studies, using the K-Means, PSO K-Means, and Robust Sparse K-Means methods, as well as three types of distance measurements: Euclidean, Manhattan, and Mahalanobis Distance. The dataset consists of 17 attributes. The evaluation was conducted using the Silhouette Score, Davies-Bouldin Index, and visualization using PCA. The results indicate that the combination of PSO K-Means with Euclidean Distance provides the best balance between quantitative performance (Silhouette Score 0.1253 and DBI 2.0521) and a more visually representative distribution of cluster members. The interpretation of the clustering results yielded three clusters: Cluster 0 (no further study) consisting of 8 employees, Cluster 1 (further study) consisting of 97 employees, and Cluster 2 (awaiting study decision) consisting of 58 employees. These findings can be utilized by institutions to design more targeted and data-driven human resource development strategies.]]>
Copyrights © 2026
