Spatial Clustering of the Village Development Index Using K-Means and DBSCAN Approaches in Machine Learning: A Case Study in Malang Regency, Indonesia

Open

A. Yudono, A.R.R.T. Hidayat, W.P. Wijayanti, F. Afrianto, A.D. Fitrianto

2026 IOP Conference Series: Earth and Environmental Science Vol. 1595 Issue 1 Conference paper Cited by 0

Abstract

This research assesses the influence of national development initiatives on rural development results, as reflected in the Village Development Index (Index Desa Membangun/IDM). The IDM is a composite measure that assesses village self-reliance on three dimensions: social, economic, and environmental resilience. Clustering tools, specifically K-Means and DBSCAN, were used to examine the distribution of social and economic resources. The findings show that K-Means regularly delivered positive, albeit relatively modest, silhouette scores, with a stable trend and improvement in 2021, indicating more fit for the dataset. DBSCAN with parameters (Iµ = 0.1, Min_Samples = 5) produced mostly negative results, suggesting poor alignment with data properties. Though more parameter modification or other techniques are needed to improve clustering quality, K-Means fared better overall than DBSCAN. © Published under licence by IOP Publishing Ltd.

Affiliations

Department of Urban and Regional Planning, Faculty of Engineering, Universitas Brawijaya, Malang, Indonesia; Department of Urban and Regional Planning, Faculty of Engineering, Universitas Gadjah Mada, Yogyakarta, Indonesia; Department of Management, BINUS Bussiness School, Bina Nusantara University, Jakarta, Indonesia