Klasterisasi Prevalensi Stunting Menggunakan K-Prototype pada Data Campuran

Marsandy, Aldwin Falah Hasan
Hayati, Memi Nor
Fauziyah, Meirinda

METIK JURNAL
Vol. 8 No. 2 (2024): METIK Jurnal

Cluster analysis is a statistical method for grouping objects based on the similar characteristics of each object. One of the algorithms used in cluster analysis is K-Prototype, which was developed to handle mixed data, namely numerical and categorical data. The validation method used to determine the optimal number of clusters in K-Prototype cluster analysis is the Elbow method. The aim of the research is to determine the optimal number of clusters and optimal cluster results on the prevalence of stunting and indicators that influence the prevalence of stunting in Indonesia in 2022. The results of the research show that the optimal number of clusters produced is 4 clusters, using the Elbow graph the WCSS (Within Cluster Sum Square) value is obtained. optimal is 65.83. Cluster 1 consists of 2 provinces, cluster 2 consists of 7 provinces, cluster 3 consists of 10 provinces, and cluster 4 consists of 15 provinces.