<![CDATA[Nutritional status among children under five remains an important public health concern, particularly in developing regions where early detection of growth problems is essential for effective intervention. Conventional nutritional assessments often rely on categorical classifications that may not fully capture variations in anthropometric characteristics among toddlers. This study aims to segment children under five based on nutritional status using the Agglomerative Hierarchical Clustering (AHC) algorithm with the Average Linkage method in the NA-IX-X District, North Labuhanbatu Regency. The study used secondary anthropometric data from 1,452 children obtained from the Aek Kota Batu Public Health Center. Quantitative variables, including body weight, height, and age, were standardized using z-score transformation prior to clustering analysis. The results show that a three-cluster configuration provides the optimal segmentation, with a Silhouette Coefficient value of 0.5154, indicating a moderate clustering structure. Cluster 1 (n = 180) shows relatively lower anthropometric measurements with an average body weight of 7.3 kg and height of 68.3 cm. Cluster 2 (n = 511) represents intermediate measurements with an average body weight of 11.5 kg and height of 87.8 cm, while Cluster 3 (n = 761) reflects higher measurements with an average body weight of 15.0 kg and height of 101.7 cm. Dendrogram analysis indicates that a cutting point at height = 1.5 produces the most interpretable cluster separation. These findings demonstrate that hierarchical clustering can support more targeted nutritional intervention strategies at the community health center level. ]]>
Copyrights © 2026
