PENERAPAN ALGORITMA K-MEANS CLUSTERING UNTUK PENGELOMPOKAN TINGKAT KESEJAHTERAAN KELUARGA UNTUK PROGRAM KARTU INDONESIA PINTAR

Abstract

Poverty in Indonesia occurs almost in various regions, one of them is the Jakarta region. Although Jakarta is a fairly large area and is one of the most developed regions, there are still some families that are still poor and unable to send their children to school, so the government issued the Kartu Indonesia Pintar program to help poor families. Although this program can really help poor families to send their children to school, sometimes there are some families that are classified as poor but do not even get the benefits of this program, but a family that can afford it get the benefits. The purpose of this paper discusses the application of data mining using the algorithm K-Means Clustering in grouping families classified as poor, simple and rich in terms of the amount of monthly income, average monthly expenditures and total assets owned, as well as the number of children who are classified according to elementary level , Junior high and high school to determine the amount of assistance that will be given based on the level. The resulting output will be in the form of a 3-table report in which each table contains data on families classified as poor, simple and rich, where this simple group can later be reviewed whether they are entitled or not to get the assistance of the Kartu Indonesia Pintar program. From the results of testing using 200 family data, the accuracy of the results of the application of K-Means Clustering was 69%.