Implementasi Algoritma Dbscan Pada Pengelompokan Desa Dan Kelurahan DI Kabupaten Jember Berdasarkan Persebaran Tingkat Kesejahteraan Sosial

Abstract

The problem of social welfare levels is still an unresolved social homework for every country. Indonesia is one of the countries that has a low level of social welfare, where information related to this can be seen from the condition profile of an area. The purpose of this research is to cluster villages and sub-districts based on social welfare levels, using the DBSCAN algorithm and Silhoutte Coefficient as an evaluation of clustering results. The results of research on the case study of clustering villages and sub-districts in Jember Regency based on the distribution of social welfare levels found that the best clustering result evaluation is eps 0.15 and minPts 5 with a total of 5 clusters and 11 noise. The results of clustering into 5 clusters obtained the first cluster (C1) as many as 68 villages and villages, the second cluster (C2) as many as 10 villages and villages, the third cluster (C3) as many as 6 villages and villages, the fourth cluster (C4) as many as 108 villages and villages, and the fifth cluster (C5) as many as 45 villages and villages, and 11 noise.