Time series prediction of earthquake input by using soft computing

Time series analysis is one of important issues in science, engineering, and so on. Up to the present statistical methods [T. Ozaki, et al. (1998)] such as AR model [T. Ozaki, et al. (1998)] and Kalman filter [S. Arimoto, (1985)] have been successfully applied, however, those statistical methods may have problems for solving highly nonlinear problems. An attempt is made to develop practical methods of nonlinear time series by introducing such soft computing techniques [L.A. Zadeh, (1965), (1973), (1993)] as chaos theory [K. Ito, (1993)], neural network [S. Chen, et al. (1989), M. Funabashi, (1992)], GMDH [A.G. Ivakhenemko, (1968), I. Hayashi, (1995)] and fuzzy modelling [(H.Nomura, et al. (1991), Y. Shi, et al. (1996)]. Using the earthquake input record obtained in Hyogo, the applicability and accuracy of the proposed methods are discussed with a comparison of those results