Clustering of Emergency Events in Surabaya Using K-Means Algorithm

Abstract

Purpose: To optimize knowledge and simplify the analysis process, the clustering method is used to group emergency
event data that occurred in the Surabaya City area. Design/methodology/approach: The clustering method used
in this research is the K-Means Clustering. Findings/result: The results of cluster 0 contain sub-districts with a high
level of vulnerability to emergencies such as Dukuh Pakis, Genteng, Gubeng, Rungkut, Sawahan, Sukomanunggal and Tambaksari sub-districts. Cluster 1 contains sub-districts with a low level of vulnerability to emergencies such as Asem Rowo, Benowo, Bulak, Bubutan, Gunung Anyar, Gayungan, Jambangan, Karang Pilang, Krembangan, Kenjeran, Lakarsantri, Mulyorejo, Pakal, Pabean Cantian, Sambikerep, Semampir, Sukolilo, Simokerto, Tegalsari, Tandes, Tenggilis Mejoyo, Wiyung, and Wonocolo. Cluster 2 contains sub districts with a moderate level of vulnerability to emergencies such as Wonokromo Sub- district. Validation of the cluster results obtained using the Silhouette coefficient is 0.610. Originality/value/state of the art: This research uses emergency incident data directly obtained from BPBD Surabaya and processed using the K-Means Clustering method.