The Power and Deflation Method based Kernel Principal Component Analysis

摘要：

Kernel principal component analysis (KPCA) a popular nonlinear feature extraction method. It gener ally uses eigen-decomposition technique to extract the principal components in feature space. But the method is injeasible for large-scale data set because of the stor age and computational problem. To overcome these disadvantages, an efficient iterative method of comput ing kernel principal components is proposed. First, the Power iteration is introduced to compute the first eigenvalue and corresponding eigenvector. Then the deflation method is repeatedly applied to achieve other higher order eigenvectors. In the process of computa tion, the kernel matrix needs not to compute and store in advance. The space and time complexity of the pro posed method is greatly reduced. The effectiveness of proposed method is validated from experimental results.

作者: Weiya Shi Dexian Zhang

作者单位: School of Information Science and Engineering Henan University of Technology,Zhengzhou,450001, China School of Information Science and Engineering Henan University of Technology,Zhengzhou, 450001, Chin

会议类型: 国际会议

会议名称: 2010 IEEE International Conference on Intelligent Computing and Intelligent Systems(2010 IEEE 智能计算与智能系统国际会议 ICIS 2010)

会议地点: 厦门

会议语种:英文

页码: 828-832

在线出版日期: 2010-10-29（万方平台首次上网日期，不代表论文的发表时间）