Recent Advances in the Use of Machine and Deep Learning for Seismic Monitoring and Early Warnings of Earthquakes

This work presents an overview of recent studies of machine learning and deep learning algorithms that are related to earthquakes. The proposed algorithms are categorized into three major groups as being for long term-earthquake hazard detection, mid-term earthquake magnitude prediction, and the short-term provision of early warnings of earthquakes. Significant research attention has been paid to deep learning and it has emphasized the classification and prediction of seismic activity for hazard detection, magnitude estimation, and the early warnings of earthquakes. The investigation of the study shows that the main uses of both machine learning and deep learning algorithms in earthquakes over recent decades have been earthquake prediction and seismic event classification. Artificial neural networks and convolutional neural networks are superior algorithms for earthquake prediction and classification when compared to others for the last decades. In future work, the balance between misidentified earthquakes and non-earthquakes should be further improved by introducing more sophisticated rules for the compilation of network discriminants using new or enhanced algorithms.