Structural recurrent neural network models for earthquake prediction

The earthquake prediction problem can be defined as given a minimum Richter magnitude scale and a specified geographic region, predicting the possibility of an earthquake in that region within a time interval. This is a long-time studied research problem but not much progress is achieved until the last decade. With the advancement of computational systems and deep learning models, significant results are achieved. In this study, we introduce novel models using the structural recurrent neural network (SRNN) that capture the spatial proximity and structural properties such as the existence of faults in regions. Experimental results are carried out using two distinct regions such as Turkey and China where the scale and earthquake zones differ greatly. SRNN models achieve better performance results compared with the baseline and the state-of-the-art models. Especially the SRNNClass(near) model, that captures the first-order spatial neighborhood and structural classification based on fault lines, results in the highest F-1 score.

Automated summary

This study introduces novel models using the structural recurrent neural network (SRNN) to capture the spatial proximity and structural properties in earthquake prediction. Experimental results in two distinct regions, Turkey and China, show that the SRNN models achieve better performance compared to baseline and state-of-the-art models. Particularly, the SRNNClass(near) model, which captures the first-order spatial neighborhood and structural classification based on fault lines, achieves the highest F-1 score.