Comparison of K-Means Algorithm and DBSCAN on Aftershock Activity in the Flores Sea: Seismic Activity 2019-2022

Abstract

This study seeks to determine whether the clustering method can be used to analyze Flores Sea earthquake activity. In this investigation, the BMKG Repo serves as the source for real earthquake vibration data collection. The stages of this research include preparing the data in CSV format and then preparing the data to eliminate useless data by identifying missing data. On the basis of the research data, it was determined that the K-Means and DBSCAN methods are used to determine the clustering method for analyzing earthquake activity. In addition, the data is depicted using a graphical Elbow method so that we can determine the number of clusters of aftershocks in the Flores Sea. The results of the visualization of aftershocks that followed earthquakes in the Flores Sea between 2019 and 2022 revealed three distinct groups of earthquake source depths: 33 to 70 kilometers, 150 to 300 kilometers, and 500 to 800 kilometers. In terms of the shilhoute index parameter, the K-Means algorithm is preferable to the DBSCAN algorithm when clustering results are used to analyze earthquake activity.