基于支持向量机的矿震危险性预测

Prediction of Mine Earthquake Risk Based on Support Vector Machine

文章介绍了支持向量机二分类的基本原理及基于此原理的三元分类问题的解决办法,分析了矿震的主要影响因素,提出了一种基于支持向量机的矿震危险性预测新方法。通过对少量训练样本的学习,能很好地建立矿震危险性等级与其影响因素之间的非线性映射关系。算例结果表明,该预测方法是可行的,预测精度较高,具有良好的工程应用前景。

This paper introduces the basic principles of dual classification by SVM and ternary classification based on this principle, an- alyzes main factors of mine earthquake, proposes a new method of forecasting mine earthquake risk based on SVM. The nonlinear rela- tionship 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 engineer- ing applications.