Optimalisasi Penentuan Klaster pada Indeks Khusus Penanganan Stunting Menggunakan Metode Agglomerative Hierarchical

Stunting is a nutrition problem that is still the main focus in developing countries, one of which is Indonesia. One of the instruments designed to measure the performance of the implementation of the stunting reduction acceleration program at the national level is the Stunting-Specific Intervention Index (Indeks Khusus Penanganan Stunting/IKPS). This study aims to group provinces in Indonesia based on a special index for handling stunting consisting of six indicators, namely health, nutrition, housing, food, education, and social protection indicators using the agglomerative hierarchical clustering method. The agglomerative hierarchical clustering method is divided into several methods, including single linkage, complete linkage, average linkage, and ward methods. This study compares the four methods with the aim of obtaining the best cluster solution in the grouping of provinces in Indonesia based on the stunting-specific intervention index. The determination of the best method in agglomerative hierarchical clustering is determined by the value of the cophenetic correlation coefficient. The results show that the average linkage method provides a better cluster solution than other methods. The cluster solution in the average linkage method produces eight clusters, including, cluster 1 consists of one province, cluster 2 consists of nine provinces, cluster 3 consists of twelve provinces, cluster 4 consists of six provinces, cluster 5 consists of one province, cluster 6 consists of one province, cluster 7 consists of three provinces, and cluster 8 consists of one province in Indonesia.