Pengelompokkan Data Kemiskinan Provinsi Jawa Barat Menggunakan Algoritma K-Means dengan Silhouette Coefficient

Abstract

Serious poverty is still one of the problems in Indonesia, especially in West Java Province. The level of underdevelopment and unemployment is still the basis for poverty. Poverty in each region is certainly different. The government needs to know which areas fall into the categories of high poverty levels and low poverty levels so that they can make solutions to set priorities for assisting. Therefore, a data mining technique is needed that can classify the poverty level of areas in West Java, namely the clustering technique with the K-Means algorithm. The purpose of this research is to classify poverty data in West Java Province so that it can be used as information to determine the right policy to distribute aid to the community from the West Java government. The results obtained based on the test, the clusters obtained were 2 clusters with cluster 0 of the high poverty level in as many as 14 regions and cluster 1 of the low poverty level in as many as 13 regions. Based on the test, the K-Means Algorithm obtains a Silhouette Coefficient of 0.576 and is included in the medium structure category. With the results of grouping poverty data, the government can channel aid more precisely.