Determining Zoning of Areas Affected by Flood Disasters in Medan City Using Silhouette Coefficient and Davies Bouldin Index Analysis

Abstract

Several sub-districts are cities that are pretty or even very vulnerable to flood disasters. Therefore, the Government must observe which areas are highly prone to flooding to anticipate safety precautions. One of the relevant methods for handling this case is k-means clustering using the Silhouette Coefficient and Davies-Bouldin Index evaluation. This research uses a quantitative approach using the Silhouette Coefficient and Davies Bouldin Index Analysis method. Based on the research that has been carried out, the results can be obtained that cluster 1 consists of the sub-districts of Medan Polonia, Medan Johor, Medan Selayang, Medan Baru, Medan Tuntungan and Medan Barat. Cluster 2 only has one member, namely the Medan Maimun sub-district; Cluster 3 consists of Medan Sunggal and Medan Belawan sub-districts, and Cluster 4 consists of Medan Marelan, Medan Helvetia and Medan Timur sub-districts. The results of the clustering evaluation that was carried out obtained a Silhouette Coefficient value of 0.505539942 and a Davies-Bouldin Index value of 0.30055. This means that the grouping carried out in this research is accurate.

Keywords

• clustering; davies bouldin indeks; k-means; silhouette