摘要
构造分类器是图像分割中重要的处理环节,论文将先验知识引入支持向量机,对支持向量机做了改善,在改善的支持向量机基础上构造实现了一个分类器,重点是将为了检验分类器的有效性,通过对UCI机器学习数据库的数据进行的实验,实验结果表明改善的支持向量机分类准确率比支传统持向量机的准确率高。
Constructing classifier is an important processing step in image segmentation .In this paper ,a prior knowl‐edge of support vectormachine is introduced ,and support vector machine is improved ,the support vector machine based on improved structure to achieve a classifier .Experiment results show that the classifier offers better classification precision than the traditional support vector machine methods through the experiment whose data from UCI machine learning database breast cancer data .
出处
《计算机与数字工程》
2015年第2期316-319,共4页
Computer & Digital Engineering
关键词
先验知识
支持向量机
医学图像分割
prior knowledge
support vector machine
medical image segmentation