Study On The Earthquake Forecast Based On Support Vector Machine And Wavelet Analysis

Earthquake is one of the most ruinous natural disasters. Its breakage is not only volcanic, but also centralized. It threatens economy , the safe of life and property. China is one of the earthquake activity fiercely countries, and its earthquake breakage is the most serious in the world. As the exigent need of government and society, taking precautions against earthquake is very significant and pressing. Earthquake prediction is the base of taking precautions against earthquake, and it is already an important problem in the seismological research at present.Wavelet is a useful tool for analyzing time-frequency of non-stable series and widely applied in many fields. It is used to analyze the series of the earthquake data for the past one hundred years in Chinese main land, and the characteristic information in different time scales is obtained. In last century, earthquake activity changes with time. It is significant that the analysis is made on the basis of a specific time scale to discuss the active and quiet periods of earthquake activity. In addition, the primary periods and the wavelet coefficients are used to analyze and discuss the trend of earthquake.Statistical learning theory (SLT) is a small-sample statistics theory. Support vector machine (SVM) is a new machine learning method based on statistical learning theory. It can process the high nonlinear problems with classification and regression. SVM not only can solve some problems, such as small-sample, over-fitting high-dimension and local minimum, but also has higher generalization (forecasting) ability than that of the artificial neural network. The time series of earthquake is not smooth, and its characteristic is non-linear, we can use SVM to study and analyze the time series of earthquake, and search after the information of the time series. It will be a effective method for exact forecasting. In this paper, SVM was used to predict the time series of the strong earthquake, and to forecast the maximum earthquake magnitude in China's mainland next year. The results show the method has a good forecasting effect.