<![CDATA[Population welfare plays a fundamental role in shaping political, economic, and social systems, as it reflects the achievement of national development goals. Welfare indicators generally encompass social, economic, and governance dimensions. This study aims to classify Indonesian provinces based on People’s Welfare Indicators using a machine learning approach, namely the Unsupervised Possibilistic Fuzzy C-Means (UPFC) method, which integrates Fuzzy C-Means and the Possibilistic Cluster Algorithm to manage uncertainty and overlapping cluster boundaries in socioeconomic data. The results identify two distinct clusters: pre-prosperous and prosperous provinces. A total of 55.88% (19 provinces) are categorized as pre-prosperous, while 15 provinces are classified as prosperous. Beyond statistical classification, the clustering pattern reveals a clear socio-regional structure. Prosperous provinces are largely concentrated in western Indonesia, particularly on Java, characterized by stronger infrastructure, economic productivity, and access to public services. In contrast, many provinces in eastern Indonesia fall into the pre-prosperous cluster, indicating persistent structural disparities. Theoretically, the findings support the uneven development perspective, demonstrating that welfare inequality in Indonesia is spatially structured rather than randomly distributed. Practically, the results provide an evidence-based foundation for targeted fiscal allocation and prioritization of basic service provision to promote more balanced and inclusive national development.]]>
Copyrights © 2026
