基于支持向量机的地震序列分类

CLASSIFICATION OF THE SEISMIC SERIES TYPE BASED ON THE SUPPORT VECTOR MACHINE

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,建立了地震序列分类模型。通过试算和分析比较得到了地震序列最佳分类模型,最佳模型的分类结果与实际地震序列分类基本一致。综合分析认为支持向量机算法无论在学习或者预测精度方面都具有很大的优越性,其获得的地震序列分类知识库可以较为准确地实现地震序列类型的分类,因此基于支持向量机理论建立的地震序列分类模型应该是可行的。

Based on the structural risk minimization principle, algorithm, in artificial intelligence field was introduced in the latest data mining method, support vector machine （SVM） this paper. Classifying models for seismic series type were established according to the linear classify algorithm and the nonlinear classify algorithm of SVM. The best classifying model about seismic series type was gained after trials and analysis, and gotten basically a good coherency between the classifying results and the actual seismic series type. The support vector machine classifying algorithm has an obvious superiority whatever on machine learning or prediction accuracy, and the classifying model for seismic series type based on the SVM theory is feasible, and it can classify the seismic serial type more accurately.