MODEL KLASTERISASI DATA PENDUDUK MENGGUNAKAN ALGORITMA K-MEANS UNTUK MENGETAHUI PRIORITAS PENERIMA BANTUAN SOSIAL DI DESA BAPINANG HULU

Central Kalimantan has a poor population of 140.04 thousand people with a poverty percentage of 5.16%. The poverty severity rate reaches a value of 0.15, with a poverty line of 506,982 IDR/Capita/Month. Bapinang Hulu Village in Central Kalimantan has around ±428,895 people in 2021. This large population makes it difficult to determine the priority of social assistance recipients, coupled with limited human resources in the village office. Data collection on social assistance recipients is still carried out based on the proposal of the RT head without proper validation, often causing social jealousy. This study aims to optimize the distribution of social assistance in Bapinang Hulu Village using the K-Means algorithm for grouping population data. The dataset consists of 246 records with 14 attributes that reflect the conditions of the head of the family in the village. The K-Means algorithm was chosen because of its ability to group data based on attribute similarities. Testing was carried out 12 times with variations in the K value to determine the optimal clustering. The results show that in the 12th test with a value of K = 13, the lowest Davies-Bouldin Index (DBI) value of 0.072 was obtained. This shows that clustering at K = 13 is optimal in terms of separation between clusters and density within clusters. Clustering helps identify community groups that need social assistance the most, provides more accurate recommendations for the priority of social assistance recipients, so that the distribution of assistance is more targeted and effective.