Analisis Pola Persebaran Hipertensi Di Sumatera Utara Dengan Menggunakan Algoritma K-Means

This is the authors' abstract. We don't add key points for this paper.

The fluctuating spread of hypertension requires accurate mapping for the allocation of medical resources. Often, regions with a consistent level of cases escape government attention compared to regions experiencing the highest case spikes in a specific year. Addressing this issue, this study aims to cluster 33 regencies/cities in North Sumatra using the K-Means Clustering algorithm to map the stability of disease spread based on 5-year historical data 2020–2024. The features used include mean prevalence X_1, growth rate X_2, and fluctuation level X_3. Data normalization was carried out using Min-Max Normalization and measured using Euclidean Distance. System testing reached convergence at the 3rd iteration, grouping regions into three clusters: High Priority C1; 15 regions, Medium Priority C2; 9 regions, and Low Priority C3; 9 regions. Quality evaluation using Silhouette Score resulted in a global average score of +0.3941 Weak Structure. The developed web-based system serves as a decision support tool for health authorities in formulating targeted medical intervention strategies.