摘要
在核函数基础上,提出了一种融合支持向量机和核主元分析的核PCA支持向量机综合集成分类方法,给出了算法实现步骤。仿真实验表明了该算法具有很好的分类性能,特别适合于消除噪声情形的模式识别问题。
Based on the kernel function, a kind of kernel PCA SVM integrated classifying method through combing the support vector machine with kernel principle component analysis is proposed, and the algorithm realizing steps are presented. The simulation experiment results show that the proposed approach has excellent classification performance. It is suitable for the pattern recognition problems requiring noise elimination.
出处
《湖南理工学院学报（自然科学版）》
CAS
2006年第4期23-26,共4页
Journal of Hunan Institute of Science and Technology(Natural Sciences)
基金
湖南省教育厅优秀青年科研项目(05B052)
关键词
核函数
核主元分析
支持向量机
分类
kernel function
kernel PCA
support vector machine
classification
