| 一种基于核主元分析的支持向量机识别方法 |
| |
| 引用本文： | 刘显贵,谢云敏,陈无畏.一种基于核主元分析的支持向量机识别方法[J].南昌大学学报(理科版),2007,31(1):49-52. |
| |
| 作者姓名： | 刘显贵 谢云敏 陈无畏 |
| |
| 作者单位： | 1. 南昌工程学院,机械与动力工程系,江西,南昌,330099
2. 合肥工业大学,机械与汽车学院,安徽,合肥,230069 |
| |
| 基金项目： | 国家自然科学基金
,
江苏省汽车重点实验室基金 |
| |
| 摘 要： | 核主元分析具有能较好地提取非线性特征的优势;支持向量机具有的非线性映射能力,且泛化能力强.它们在分类与识别中应用时都各有自己的优点,结合核主元分析和支持向量机的特点,提出一种基于核主元分析的支持向量机识别方法,用该方法分别对 ORL 人脸库和iris数据集中的数据进行分类与识别,结果表明:如果根据设计好的核函数的参数,可以得到极高的识别率.
|
| 关 键 词： | 核主元分析 支持向量机 分类 识别 核主元分析 支持向量机 识别方法 Principal Component Analysis Kernel Based Method Support Vector Machine 识别率 参数 核函数 设计 结果 数据集 iris 人脸 结合 应用 分类与识别 泛化能力 |
| 文章编号： | 1006-0464（2007）01-0049-04 |
| 收稿时间： | 2006-05-20 |
| 修稿时间： | 2006-05-20 |
Research of Support Vector Machine Classified Method Based on Kernel Principal Component Analysis |
| |
| LIU Xian-gui,XIE un-ming,CHEN Wu-wei.Research of Support Vector Machine Classified Method Based on Kernel Principal Component Analysis[J].Journal of Nanchang University(Natural Science),2007,31(1):49-52. |
| |
| Authors: | LIU Xian-gui XIE un-ming CHEN Wu-wei |
| |
| Affiliation: | 1. Departmemt of Mechanical and Power Engineering, Nanehang Institute of Technology, Nanehang 330099, China ; 2. School of Mechanical and Automotive Engineering, Hefei University of Technology, Hefei 230069 ,China |
| |
| Abstract: | Kernel Principal Component analysis(KPCA) has the advantage of extracting nonlinear features.Nonlinear mapping and generalization are the strong capabilities of Support Vector Machine(SVM).they have own advantages when they each is applied into classification and identifition.By integrating the characteristics of KPCA and SVM,a SVM recognized method based on KPCA is put forward,and is carried on the recognition to the data of the ORL person face database and the iris data concentration,the result shows that the recognized accuracy can reach 98.6% according to designed the parameters. |
| |
| Keywords: | kernel principal component analysis S V M classification identifition |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
