摘要
提出一种基于一类支持向量机(one-class SVM)的贝叶斯分类算法,该算法用一类SVM对类条件概率密度进行估计以构造贝叶斯分类器.证明采用高斯核的一类SVM,其解可以归一化为密度函数,并把该密度函数看作类条件概率密度的平滑估计,构造贝叶斯分类器.实际数据集上的实验结果表明,提出的分类算法测试准确率高于简单贝叶斯分类器与贝叶斯网络分类器,不低于传统二类SVM;比传统二类SVM需要计算的核矩阵规模更小,训练时间更短.
A Bayesian classification algorithm based on one-class SVM is presented. It constructs the Bayesian classifier using the classes' conditional density estimated by one-class SVM. It is proven that the solution of one-class SVM using the Gaussian kernel can be normalized as an estimate of probability density, and can be used to obtain the Bayesian classifier. Experimental results showed that the proposed classifier outperformed NaiveBayes and BayesNet in terms of prediction accuracy, comparable to traditional two-class SVM. The size of kernel matrix of the new algorithm is less than that of the traditional two-class SVM, which lead to less training time for the new classifier.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2006年第2期143-146,共4页
Transactions of Beijing Institute of Technology
基金
国家"九七三"计划项目(G1998030414)
关键词
贝叶斯分类
支持向量机
概率密度估计
Bayesian classification
support vector machine, probability density estimate