Classification of Natural Disaster Prone Areas in Indonesia using K-Means

Disaster caused by both nature and human factors has resulted in the occurrence of human casualties, environmental damage, property loss, and psychological impact. The study aims to classify disaster prone areas in Indonesia using K-means clustering method implemented in rapid miner tools. The data are collected from the Central Bureau of Statistics about the number of villages that considered as natural disaster-prone by province in Indonesia in years [2008][2009][2010][2011][2012][2013][2014]. The sample data are 34 provinces in Indonesia with 3 natural disasters commonly happen i.e. namely: Flood, Earthquake and Landslide. The final outcomes of the study were: (1) 4 provinces classified as High with cluster center 1363.333 (flood), 528.25 (earthquake) and 949.583 (landslide); 14 provinces classified as Medium with cluster center 142.619 (flood), 96.071 (earthquake) and 72.048 (landslide); and 16 provinces classified as Low with cluster center 507.396 (flood), 57.604 (earthquake) and 177.479 (landslide). This work can further provide input to the Indonesia government through mapping of disaster prone areas especially 4 provinces with very high natural disasters such as Aceh, West Java, Central Java and East Java.