Random Forest Analysis for Predicting the Probability of Earthquake in Indonesia

Abstract

This research focuses on identifying risk zones by applying the Random Forest algorithm to predict the probability of earthquakes in Indonesia. The selection of this algorithm is based on its capacity to process voluminous, intricate, and non-linear data, which is frequently encountered in the context of seismic studies. In this study, a predictive model is constructed using historical earthquake data and geographic coordinates. The primary objective is to evaluate the effectiveness of the Random Forest algorithm in predicting earthquake probabilities across different regions of Indonesia. The analysis results indicate that the highest likelihood of earthquakes occurs in Maluku at 24.77%, followed by Nusa Tenggara at 18.34% and Sulawesi at 18.68%. The Random Forest algorithm achieved an accuracy rate of 90.78% in the prediction model, demonstrating its effectiveness in forecasting earthquake probabilities. These findings are expected to provide valuable insights for the government and stakeholders to develop improved disaster mitigation strategies in Indonesia. Furthermore, the methods used in this study can be applied to predict the probabilities of various types of natural disasters across different regions. on using larger datasets and examining the specific regions from which the data is collected.