Klasifikasi Kedalaman Kejadian Gempa Menggunakan Algoritma K-Means Clustering: Studi Kasus Kejadian Gempa Di Sulawesi

Abstract

Sulawesi region is one of the region that have complex geologic conditions so that disasters caused by large scale earthquake frequently occur in these region. Depth and magnitude attribute of the earthquake that cause the disasters are investigation using machine learning technique. Longitude, latitude, magnitude, depth attributes are used to depth cluster of the earthquake events in 1970-2022 period. The cluster number have been optimized by Elbow method, and validated by Davies-Bouldin index (DBI). The result is shown that the three cluster is the best cluster than the others, and its Davies-Boludin index is 0.397. Depth of the fist cluster is less than equal to 120 km (shallow earthquake), the second cluster is among 120 km and 350 km (intermediate earthquake), and the third cluster is greater than 350 km (deep earthquake). The cluster visualizations of the earthquakes are revealed that shallow earthquakes with above 5 SR are frequently occurred in shallow depth. Based on results, Sulawesi Region is vulnerable to earthquake hazard, and K-Mean clustering algorithm is successfully to cluster of earthquake depth.