Radial basis function neural networks to foresee aftershocks in seismic sequences related to large earthquakes

Abstract

Radial Basis Function Neural Network are known in scientific literature for their abilities in function approximation. Above all, this particular kind of Artificial Neural Network is applied to time series forecasting in non-linear problems, where estimation of future samples starting from already detected quantities is very hardly. In this paper Radial Basis Function Neural Network was implemented in order to predict the trend of n(t) for aftershocks temporal series, that is the numerical series of daily-earthquake’s number occurred after a great earthquake with magnitude M > 7.0 Richter. In particular we implemented the RBF-NN for the Colfiorito seismic sequence. The seismic sequences considered in this work are obtained following criteria already known in scientific literature. Results of proposed approach are very encouraging.