EUVIMEDEUVIMED entdecken
Open Access · CC BY-SA 4.0

Comparison of K-Means and K-Medoids for PKH Priority Clustering

M. Ahsanul Fikri, Ifnu Dwi Wisma Prastya, Alif Yuanita Kartini

Journal of Applied Informatics and Computing · 16.04.2026 · 10 (2) · 1347-1358

DOI: 10.30871/jaic.v10i2.12267 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Applied Informatics and Computing, online erschienen am 16.04.2026. Verfasst von M. Ahsanul Fikri und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The Family Hope Program (PKH) is a government policy aimed at poverty alleviation and protecting vulnerable groups. Differences in the number and composition of PKH recipients between sub-districts indicate variations in the level of social vulnerability and potential regional inequality that require objective mapping. This study aims to identify the concentration of socially vulnerable groups and inequality between sub-districts in Bojonegoro Regency through a data-based cluster approach. The data used comes from the Bojonegoro Regency Social Service which includes PKH recipient components, namely early childhood, SD, SMP, SMA, people with severe disabilities, the elderly, and pregnant women. This study compares several clustering scenarios including the K-Means and K-Medoids algorithms, variations in the number of clusters (k = 2 and k = 3), Euclidean and Manhattan distance functions, and data conditions without normalization and with Min–Max normalization. The determination of the number of clusters is done using the Elbow method, while cluster quality is evaluated using the Silhouette Score. The results of the study show that the best scenario was obtained in K-Medoids with Euclidean distance and Min–Max normalization at k = 3 with a Silhouette Score value of 0,8115. The clustering results divided the sub-districts into three categories, namely high clusters (many) with 10 sub-districts, medium with 13 sub-districts, and low (few) with 5 sub-districts. These findings can be used as a basis for determining the priority of policy interventions, especially in high clusters, so that the distribution of PKH becomes more targeted and is able to reduce social inequality between regions.

Zitieren

APA

Fikri, M. A., Prastya, I. D. W., & Kartini, A. Y. (2026). Comparison of K-Means and K-Medoids for PKH Priority Clustering. Journal of Applied Informatics and Computing, 10(2), 1347-1358. https://doi.org/10.30871/jaic.v10i2.12267

Vancouver

Fikri MA, Prastya IDW, Kartini AY. Comparison of K-Means and K-Medoids for PKH Priority Clustering. Journal of Applied Informatics and Computing. 2026;10(2):1347-1358. doi:10.30871/jaic.v10i2.12267

BibTeX
@article{Fikri2026,
  author = {Fikri, M. Ahsanul and Prastya, Ifnu Dwi Wisma and Kartini, Alif Yuanita},
  title = {Comparison of K-Means and K-Medoids for PKH Priority Clustering},
  journal = {Journal of Applied Informatics and Computing},
  year = {2026},
  volume = {10},
  number = {2},
  pages = {1347--1358},
  doi = {10.30871/jaic.v10i2.12267}
}
Kostenlos starten →