Usage of multiple RTL features for Earthquake prediction

TL;DR

This paper presents a machine learning approach utilizing multiple RTL features to improve earthquake prediction accuracy, achieving high precision and recall on historical Japanese earthquake data.

Contribution

It introduces a novel machine learning model that combines various RTL features for more accurate earthquake prediction over traditional expert forecasts.

Findings

Precision up to ~0.95 on historical data

Recall up to ~0.98 on historical data

Effective prediction over 30-180 days range

Abstract

We construct a classification model that predicts if an earthquake with the magnitude above a threshold will take place at a given location in a time range 30-180 days from a given moment of time. A common approach is to use expert forecasts based on features like Region-Time-Length (RTL) characteristics. The proposed approach uses machine learning on top of multiple RTL features to take into account effects at various scales and to improve prediction accuracy. For historical data about Japan earthquakes 1992-2005 and predictions at locations given in this database the best model has precision up to ~ 0.95 and recall up to ~ 0.98.