Penerapan Algoritma Fuzzy C-Means Untuk Pengelompokan Data Penduduk Miskin Di Indonesia Berdasarkan Kabupaten dan Kota

Abstract

Poverty is a central issue for every country in the world, especially for developing countries. In Indonesia, poverty has become a phenomenon and fact, one of the problems that has not been resolved until now by both the central government and the regional government. Difficulties in determining which regions experience the highest and normal poverty levels and areas with low poverty levels, a method is needed to help this problem. One of them is data mining using clustering techniques. Clustering is a method used to group data based on the similarity of the data. In analyzing partition-based clusters, the Fuzzy C-Means (FCM) algorithm is an algorithm that has been widely used to solve data clustering problems. The variable in this study is the number of poor people based on regencies and cities in Indonesia from 2015 to 2017. These variables are used to obtain categories from each cluster formed. From the results of the analysis carried out, it can be concluded that three regencies and cities can be grouped: cluster 1 consisting of 39 districts / cities with high poverty levels, cluster 2 consisting of 368 districts / cities with low poverty level categories, and cluster 3 having members 107 districts / cities with moderate poverty levels. And from the DBI value obtained in the FCM algorithm is equal to 0.524 which means that with this value the cluster in the algorithm that is formed can be said to be good or optimal because the DBI value is close to 0.

Keywords: Clustering, Poverty Data, Fuzzy C-Means, Davies-Bouldin Index