摘要
该文以多个支持向量机分类器的输出向量为基础,运用决策模板来估计各个分类器的精度,然后使用常见的融合规则实现融合算法,并将其运用到蛋白质结构类分类当中。实验表明:该算法可有效提高分类精度,因此具有一定的应用价值。
In order to rais e the performance of multiple classifiers fusion,a new fusion algorithm based on the output vectors of support vector machine and the accuracy of the classifier based on decision template is presented in this paper.Protein structure class classification experiment reveals that this algorithm can effectively enhance t he accuracy of recognition,so it may have application value.
出处
《计算机工程与应用》
CSCD
北大核心
2004年第19期66-67,共2页
Computer Engineering and Applications
基金
西北工业大学研究生创业种子基金(编号:Z20030048)
关键词
分类器融合
支持向量机
蛋白质结构类
classif iers fusion,support vector machine,protein structure class
