Prediction of Mine Earthquake Risk Based on Support Vector Machine

This paper introduces the basic principles of dual classification by SVM and ternary classification based on this principle,analyzes main factors of mine earthquake,proposes a new method of forecasting mine earthquake risk based on SVM.The nonlinear relationship between the mine earthquake risk level and its influencing factors can be well established through the study of a small amount of training samples.The numerical results show that the prediction method is feasible and accurate,it's with good prospects for engineering applications.