Pengelompokan Kabupaten/Kota Di Pulau Kalimantan Dengan Fuzzy C-Means Berdasarkan Indikator Kemiskinan

Abstract

Cluster analysis is a branch of statistical science that is used to grouping data that have similar characteristics between each other. The grouping method used in this research is Fuzzy C-Means. Fuzzy C-Means method is one of the grouping methods developed from the C-Means method by applying the properties of fuzzy sets. With the existence of each data is determined by the degree of membership. This method is applied to data from 56 districts/cities on Borneo based on poverty indicators with variables namely the percentage of average length of schooling, life expectancy, percentage of the poor, percentage of open unemployment rate, percentage of households with proper sanitation, and percentage of households with proper drinking water. This study aims to obtain the results of grouping districts/cities on Borneo based on poverty indicators and to obtain optimal cluster results based on three validity indices, namely Connectivity, Dunn, and Silhoutte values. Based on the results of the study, it was found that there were 2 optimal clusters, namely the first cluster consisted of 36 regencies/cities while the second cluster consisted of 20 regencies/cities.