<![CDATA[Natural disasters such as floods, earthquakes, and landslides are recurring threats in Cirebon City, West Java. This study aims to classify disaster-prone areas using the K-Means algorithm based on 1,144 incident data from Open Data Jabar. The data were grouped into three clusters, namely safe, moderate, and dangerous. Cluster quality was evaluated using the Silhouette Score and Elbow Method. The results of this study show that the model without normalization produced a score of 0.6804, reflecting good cluster separation. Conversely, the application of MinMaxScaler normalization significantly reduced the model's performance, with a score of 0.3900. The main contribution of this study is to show that data normalization can disrupt the natural pattern of risk distribution, thereby reducing the quality of clustering. Therefore, the selection of pre-processing techniques needs to be adjusted to the characteristics of local data. It is hoped that this study can be the basis for the development of a more adaptive and data-driven disaster mitigation decision support system.]]>
Copyrights © 2025
