摘要
支持向量机方法基于结构风险最小化原理，克服了常规统计方法的局限性，能够在有限的样本集基础上兼顾模型的通用性和推广性，预测精度更高。因此，利用支持向量机方法对地震储层厚度进行了预测，预测结果与实际情况吻合。与BP神经网络预测结果进行对比分析的结果表明，采用支持向量机方法预测的效果较好，是一种值得推广的方法。
Support vector machine method is established on the basis of the .principle of structural risk minimization. It overcomes the limitation of conventional statistics method and can combine model generalization and popularization on the basis of finite sample set, with higher predication accuracy. Using this method, reservoir thickness is predicted by seismic interpretation, and results are accordant with the fact. Comparison with predicted results by Bp neural network shows that, effect of this prediction method is better, being worth to be popularized.
出处
《大庆石油地质与开发》
CAS
CSCD
北大核心
2009年第1期34-36,共3页
Petroleum Geology & Oilfield Development in Daqing
基金
国家“863”研究项目（2007AA060500）资助.
关键词
支持向量机
结构风险最小化
地震属性
储层厚度
神经网络
support vector machine
risk minimum of a structure
seismic attribute
reservoir thickness
neural network
