Clustering Earthquakes in West Java Using Machine Learning Algorithm

It is important to understand seismicity in West Java, Indonesia, which lies within the Pacific Ring of Fire and is prone to seismic events. We analyze data from earthquakes that occurred between September 2022 and September 2023. We use the K-medoids clustering method to organize the data into clusters with medoids as central representatives. This provides deeper insights into the patterns of seismic activity. As a result, we identified
$\mathrm{K}=15$
as the optimal number of clusters through the silhouette and Dunn index. This segmentation provided a very effective solution. In particular, the K-medoids revealed significant seismic clusters in West Java, which has implications for earthquake monitoring and hazard assessment in the region.