Implementation of the Random Forest Algorithm for Magnitude Prediction in the Matano Fault Indonesia

Studying earthquake prediction is an interesting academic area. Earthquakes are classified as natural disasters that have the potential to inflict significant damage.  The magnitude of an earthquake provides substantial benefits, both immediately after the event and in the future, for risk assessment and mitigation. This study employs the Random Forest algorithm to predict the magnitude of earthquakes occurring on the Matano Fault in Sulawesi, Indonesia. The prediction is derived from the historical seismic data collected between 1923 and 2023, obtained from the BMKG and USGS Catalogs. The area of interest is situated along the Matano Fault in Sulawesi, Indonesia, with coordinates ranging from 2.99°S to 1.66°S and from 120.50°W to 122.47°W. The dataset comprises six attributes and is split into training and testing sets at a ratio of 70% and 30%, respectively. The variables employed in this investigation encompass origin time, latitude, longitude, magnitude, and magnitude type.  The investigation yields an RMSE score of 0.1929. Overall, the prediction model has outstanding performance, with a high degree of accuracy in predicting values that quite match the actual values. Additionally, the Root Mean Square Error (RMSE) value is 0.1929, indicating a low level of error in the predictions. this work attempts to propose an alternative approach, characterized by a straightforward and practical technique, to solve problems in the geophysics field, for instance in the area of earthquake prediction.