Improving Regional Clustering Based on Tuberculosis Cases using the K-Means Algorithm of the Cirebon City Health Office

Tuberculosis (TB) is a highly infectious disease prevalent in Indonesia, including Cirebon City. This study utilizes the K-Means algorithm to optimize the clustering of areas based on TB case data from the Cirebon Health Office. By analyzing the number of cases, population density, and other factors, the study aims to identify regional clusters with similar TB case characteristics. The research employed Rapid Miner software and the Knowledge Discovery Database (KDD) methodology. The K-Means analysis categorized the study area into two clusters. Cluster_0, representing 20 areas, had lower TB risk, characterized by higher population density, smaller geographic size, and fewer TB cases. Cluster_1, representing two areas, exhibited higher TB risk, marked by lower population density, larger area, and more TB cases. The clustering quality was evaluated using the Davies-Bouldin Index (DBI), which yielded an optimal value of 0.189 at K=2K = 2. Additionally, the Avg within Centroid Performance Vector Analysis supported the clustering validity the clusters with value of 19851032.925.The results demonstrate that this clustering approach effectively identifies TB risk areas, aiding targeted interventions. The findings provide the Cirebon Health Office with a framework for better resource allocation, focusing intensive programs in high-risk regions and preventive measures in low-risk areas.