支持向量机在砂土液化预测中的应用研究  

Application of the support vector machine in prediction of sand liquefaction

作　　者：陈荣淋[1] 林建华[1] 黄群贤[1] 

机构地区：[1]华侨大学土木工程学院,福建泉州362021

出　　处：《中国地质灾害与防治学报》2005年第2期15-18,23,共5页The Chinese Journal of Geological Hazard and Control

基　　金：福建省自然科学基金资助项目(D020014)

摘　　要：介绍了人工智能领域最新的基于结构风险最小化原理的数据挖掘算法———支持向量机算法。根据支持向量机线性分类和可以具有不同核函数的非线性分类两种算法,建立了砂土液化预测模型,并且运用Matlab语言编写了程序。通过试算和分析比较得到了最佳模型,最佳模型的预测结果与实际液化情况基本上一致。认为支持向量机算法无论在学习或者预测精度方面都有很大的优越性,而基于支持向量机理论建立的砂土液化预测模型是可行的,且可以较为准确地实现砂土液化的预测。Based on the structural risk minimization principle, the latest data mining method, support vector machine(SVM) algorithm, in artificial intelligence field was introduced in this paper. Models for sand liquefaction prediction were established according to the linear classify algorithm and the nonlinear classify algorithm of SVM, and a program was compiled in language Matlab. The best model was gained after trials and analysis, and gotten a good coherency between the prediction results and the actual liquefaction. The support vector machine algorithm has an obvious superiority whatever on machine learning or prediction accuracy, and the model for sand liquefaction prediction based on the SVM theory is feasible, and it can forecast the sand liquefaction more accurately.

关 键 词：支持向量机 砂土液化 预测模型 线性分类算法 非线性分类算法 

分 类 号：P642.11[天文地球—工程地质学] P642.2[天文地球—地质矿产勘探]