为了实现地层识别,根据测井数据建立SVM地层识别模型,并且该模型以径向基函数为核函数。通过时SVM分类器中惩罚参数优选,结合实际需要并加入一个判断函数,提高了SVM分类器的分类正确率。试验表明:SVM分类器应用在地层识别中效果良好,而且具有广阔的应用前景。
In order to recognize stratums,a new Support Vector Machine Model(SVMM) is proposed and implemented with term of well logging data,and RBF is adopted as kernel function in SVMM.By means of the parameter C of the model adjusted optimally,and a judge-function introduced practically,the classification accuracy of SVMM is improved well.The experimental resuh shows that the SVM classifier can be applied in stratum recognition effectively,and be used widely in the future.