Implementation of K-Means Method for Clustering Natural Disaster Priority Areas in North Sumatra

Abstract

Indonesia is highly vulnerable to natural disasters due to its geographical and geological conditions, including North Sumatra Province, which experiences diverse disaster risks. This study aims to identify and group natural disaster priority areas in North Sumatra using a data-driven approach. The research employs a quantitative method by applying the K-Means clustering algorithm to group regions based on disaster occurrence and impact characteristics. Secondary data of natural disaster events in 2024 were processed through data preprocessing stages, including data cleaning and normalization. The clustering process was conducted using five clusters representing very low, low, medium, high, and very high disaster priority levels. The quality of clustering results was evaluated using the Davies–Bouldin Index (DBI). The results show that most regions fall into the very low and low priority clusters, while several regions are classified into higher priority levels. Deli Serdang Regency is identified as the very high priority area, indicating the highest disaster impact among regions. The DBI value of 1.102688 indicates that the formed clusters have reasonable compactness and separation. This study concludes that the K-Means method is effective for grouping disaster-prone areas and can support data-driven decision-making in disaster mitigation and planning in North Sumatra Province.