Comparison of B-Value Predictions as Earthquake Precursors using Extreme Learning Machine and Deep Learning

Abstract

Knowledge of earthquake predictions is very important, especially to identify patterns of occurrence of earthquakes based on data obtained from the Meteorology and Geophysics Agency (MGA). This paper proposes an earthquake prediction system, in the form of predicting the b-value as a parameter that indicates the precursor to earthquakes. A precursor is something that precedes or is thought to indicate the appearance of something, in this case, an earthquake. The paper considers two methods which are Extreme Learning Machine and Deep Learning. The simulation results show, in the training process, Deep Learning produces better b-value prediction performance as an earthquake precursor compared to Extreme Learning Machine. Meanwhile, in the testing process, the Extreme Learning Machine produces a slightly better b-value prediction performance as an earthquake precursor compared to Deep Learning. Both in the training process and in the testing process, in solving the case of predicting b-values as earthquake precursors, deep learning is more superior.