Seismic Data Forecasting: A Sequence Prediction or a Sequence Recognition Task

Abstract

In this paper, we have tried to predict earthquake events in a cluster of seismic data on pacific ring of fire, using multivariate adaptive regression splines (MARS). The model is employed as either a predictor for a sequence prediction task, or a binary classifier for a sequence recognition problem, which could alternatively help to predict an event. Here, we explain that sequence prediction/recognition, as two aspects of sequence learning, are not the same in general. We show that while both these approaches are plausible for earthquake prediction, the forecasting results indicate that MARS as a binary classifier outperforms the predictor MARS. The results clearly show how it is important to challenge a single earthquake forecasting problem from an appropriate point of view.