On Back-Propagation Network to Early Judgment of Seismic Sequences

摘要

The early predictions of earthquake sequence types are studied using BP Network. Back-Propagation network is a feedforward neural network practiced by back propagation algorithm, is one of neural network modes applied widely. It not only can approximate any continuous function, has the strong nonlinear mapping ability, but also has a strong robustness, memory capacity and self-learning ability. On the basis of Ms ≥ 5.0 earthquake sequence materials in our country since 1970, It is effective for early predictions of earthquake sequence types that we divides the data in 5 time scales according to 1-7 days after the earthquake.