The Implementation of Spatial Model with K-Means Clustering Method to Cluster Flood Affected Areas in Bone Regency

Abstract

This research is an applied research that aims to determine the clusters of areas affected by floods in bone regency. This study user the K-means method in clustering flood data in 2020 and 2021. The data is grouped based on the number of families affected and the resulting damage. In determining the number of clusters to be used, this study validated 3, 4 and 5 clusters using the davis bouldin index (DBI) validation in determining the best cluster. The results of this validation resulted in the best number of clusters, namely k=4 for 2020 data with a minimum DBI of 0,73 and k=4 for 2021 data with a minimum DBI of 0,44. After clustering, the cluster number for data for 2020 from the firs to fourth cluster member are 34, 7, 2 and 3 sequentially, while the data for years 2021  sequentially are 3, 22, 3 and 1. Then the cluster results are displayed in a spatial form that is created using ArcGIS.