IMPLEMENTASI ALGORITMA K-MEANS UNTUK CLUSTERING DATA PENYANDANG MASALAH KESEJAHTERAAN SOSIAL DI WILAYAH SUMATERA UTARA

Penyandang Masalah Kesejahteraan Sosial (PMKS)are community groups facing difficulties in meeting basic life needs due to factors such as poverty, disability, drug dependence, and disasters. In North Sumatra Province, the distribution of PMKS varies significantly across regions, requiring targeted and region-specific social policies. This study aims to cluster 33 districts/cities based on the number and types of PMKS in 2022 using the K-Means algorithm. This method was selected for its effectiveness in uncovering patterns from complex datasets. The clustering process involved multiple iterations until convergence, and the results were validated using RapidMiner. The findings reveal three clusters: Cluster 1 (two regions) with a high level of PMKS, Cluster 2 (six regions) with a moderate level, and Cluster 3 (25 regions) with a low level. Regions with high PMKS levels, such as Batu Bara and Karo, show dominance in categories like extreme poverty, disability, and socially vulnerable women. The results provide a clearer picture of social welfare conditions across the region and serve as a valuable reference for designing more focused and efficient social welfare policies tailored to regional needs.