<![CDATA[The rapid population growth in Indonesia poses significant challenges to public welfare, economic stability, and sustainable development. The Family Planning program aims to regulate population growth through various contraceptive methods; however, participation rates often differ across regions. Understanding these variations is crucial for designing targeted interventions. This study investigates how different distance measures in the K-Means clustering algorithm affect the segmentation quality of KB participants in Kalirejo Village, Lawang District. Eight distance metrics—Euclidean, Manhattan, Minkowski, Chebyshev, Mahalanobis, Bray-Curtis, Canberra, and Cosine—were compared using standardized data from the local BKKBN office (January–September). Cluster validity was evaluated using the Silhouette Coefficient across k=2–10. Results show that the Manhattan distance with k=2 achieved the best clustering quality (SC = 0.7191), effectively distinguishing participant groups by contraceptive method preference. The study highlights the importance of selecting suitable distance measures to improve data-driven policy and decision-making in family planning management.]]>
Copyrights © 2026
