A Study of Grouping of Earthquake Damage from Magnitude Scale in Lombok Using K-Means Modeling

This study aims to group earthquake damage from its magnitude scale and visualize it on a geographical map. The magnitude of the earthquake was grouped using the K-means model. It is one of the most popular and effective clustering models in grouping data, such as earthquake data. The dataset used in this study is earthquake data for the last ten years on Lombok Island. The optimal number of clusters was used which is 2 in this case, based on the highest Silhouette score of 0.930. The highest Silhouette score shows the optimal number of clusters. The cluster on the geographical map shows most earthquakes' distribution in Northen Lombok Island with cluster 1 consisting of 145 earthquakes, while cluster 2 consists of 3 earthquakes. In addition, the earthquake's damage based on its magnitude scale, there were four different kinds of earthquake damage: slight, limited, minor, and severe damage that have occurred for the last ten years in Lombok Island. Minor and Slight damages were dominant, respectively. However, severe damage occurred in the northern part of Lombok Island due to an earthquake in 2018.