Assessing the performance of K-means and DBSCAN clustering methods in tuberculosis mapping

Abstract

Tuberculosis (TB) remains a significant public health challenge, particularly in densely populated and under-resourced areas such as West Java, Indonesia. This study aimed to map and analyze the spatial distribution of TB prevalence by clustering regions based on variables including sanitation quality, population density, and TB rates. Secondary data from official sources were utilized, and clustering methods such as K-means and DBSCAN were employed to group districts and cities into distinct clusters. The K-means method identified five clusters, while DBSCAN formed four clusters with some noise. Performance evaluation using the silhouette index indicated that K-means outperformed DBSCAN. The clustering results revealed that regions with poor sanitation and high TB prevalence require prioritized public health interventions. This analysis underscores the potential of clustering methods to enhance public health planning by identifying areas in critical need of targeted TB control strategies, optimizing resource allocation, and supporting evidence-based decision-making.