Implementation of the K-means Clustering Method on Stunting Case in Indonesia

Abstract

Method K - means clustering is one technique that will partition the data into groups, so that the data which have the samecharacteristics are grouped into the same group. Clustering can be done on various types of data including clustering in order todetermine the province which is the priority for stunting reduction. This study aims to classify provinces based on the prevalenceof stunting for infants 0-59 months, exclusive breastfeeding, weigh 4 times, adequacy of energy and protein using the K-Meansclustering method. The results showed that the optimal cluster formed was 4 clusters, in which group 1 was acluster whose province had good nutritional status with an alow prevalence of stunting with a percentage of exclusivebreastfeeding, weighing 4 times, adequate energy and high protein. While cluster 4 is a cluster that needs priority attention sincethe average value of stunting high-value percentage of energy adequacy low protein,although it has a value of percentage of ASIexclusive and toddlers weighing 4 times high enough that consists of Nusa Tenggara Barat, Nusa Tenggara East, Aceh, South Sulawesi, West Sulawesi and Banten. Provinces with stunting values, exclusive breastfeeding, weighing toddlersmore than 4 times, adequacy of energy and protein that are the same height tend to group together, while provinces with lowscores also group themselves.