k-Means and GIS for Mapping Natural Disaster Prone Areas in Indonesia

Abstract

The number of natural disasters in Indonesia is very high frequency. However, the data collected based on natural disasters has complex structures. One of the efforts to make prevention design is grouping the areas of natural disasters based on their similarities. The proposed methods are k-means to cluster areas and Geographical Information System (GIS) to improve visualization of yielded clusters. This result showed that the best cluster was seven clusters based on root mean square standard deviation (RMSD). Although k-means obtained the best number of clusters, however, it was difficult to present the clusters of natural disaster areas in a map. Therefore, the GIS method can be a useful tool to improve the visualization of k-means.