<![CDATA[This study segments Indonesian provinces based on district road stability characteristics using K-Means and Hierarchical Clustering approaches. We analyzed district road stability data from 34 provinces during 2016-2023, including total road length, stable road conditions, and unstable road conditions. Data preprocessing included cleaning, normalization using min-max scaling, and feature selection. Results showed optimal clustering with k=4, achieving silhouette coefficient of 0.647 for K-Means and 0.623 for Hierarchical Clustering. Four distinct provincial clusters emerged: Optimal Infrastructure Provinces (>80% stability), Developing Infrastructure Provinces (60-80% stability), Infrastructure Challenge Provinces (<60% stability with extensive networks), and Limited Infrastructure Provinces (small networks with variable stability). The Adjusted Rand Index of 0.78 demonstrated high agreement between methods. This segmentation provides evidence-based insights for targeted infrastructure policy formulation in Indonesia.]]>
Copyrights © 2025
