Penggunaan Index Calinski-Harabasz Pada Clustering K–medoids Algorithm Untuk Penggolongan Kabupaten/ Kota Di Provinsi Jawa Tengah Berdasarkan Karakteristik Penduduk Miskin

Poverty is a problem that occurs almost every year. The government is trying hard to reduce poverty because the poverty rate is a measure of the success of a region. Clustering analysis can assist the government in providing targeted assistance. The k-medoids method is a non-hierarchical clustering method for classifying n objects into k clusters based on similar characteristics. This clustering algorithm uses the medoid as its cluster center. The k-medoids method used to overcome the problem of outliers and determine the optimal number of clusters using cluster validation. This research used Index Calinksi-Harabasz. Based on the result of the clustering, the optimal cluster was obtained k=2 using k-medoids method and cluster validation Index Calinksi-Harabasz of 31,53654. Cluster 1 consists of 22 districts or cities and in cluster 2 consists of 13 districts or cities.