[1]邓 貌,陈 旭,陈天翔,等.采用核聚类分析的KPCA改进算法[J].智能系统学报,2010,5(03):221-226.
 DENG Mao,CHEN Xu,CHEN Tian-xiang,et al.mproved kernel principal component analysis based ona clustering algorithm[J].CAAI Transactions on Intelligent Systems,2010,5(03):221-226.
点击复制

采用核聚类分析的KPCA改进算法(/HTML)
分享到:

《智能系统学报》[ISSN:1673-4785/CN:23-1538/TP]

卷:
第5卷
期数:
2010年03期
页码:
221-226
栏目:
出版日期:
2010-06-25

文章信息/Info

Title:
mproved kernel principal component analysis based ona clustering algorithm
文章编号:
1673-4785(2010)03-0221-06
作者:
邓 貌1陈 旭1陈天翔2王徽蓉1鲁华祥1
1.中国科学院 半导体研究所,北京 100083;
2.厦门理工学院 电子与电气工程系,福建 厦门 361005
Author(s):
DENG Mao1 CHEN Xu1 CHEN Tian-xiang2 WANG Hui-rong1 LU Hua-xiang1
1.Institute of Semiconductors, Chinese Academy Sciences, Beijing 100083, China;
2.Department of Electronic and Electrical Engineering, Xiamen University of Technology, Xiamen 361005, China
关键词:
核主分量分析核聚类子集划分协方差矩阵特征向量
Keywords:
KPCA kernel clustering partition of training data set covariance matrix eigenvector
分类号:
TP18
文献标志码:
A
摘要:
为了解决核主分量分析方法处理大训练样本集时计算代价巨大的问题,在采用子集划分的KPCA算法基础上,提出采用核聚类划分子集,并用每个子集的协方差矩阵的特征值累积贡献率作为标准来选取相应的特征向量.分别在人工和实际数据集上测试,实验结果显示在同一累积贡献率和给定子集个数的条件下,采用核聚类划分子集总能得到较小尺寸的核矩阵,而核矩阵尺寸的减小有助于改善测试样本的特征提取速度以及降低特征分解核矩阵的时间复杂度.
Abstract:
To overcome the computational problems of the standard kernel principal component analysis (KPCA) algorithm, the authors proposed a new method for eigenvector selection by evaluating the cumulative contribution rate of the eigenvalues of the covariance matrix. In addition, a new way to partition the training data set based on kernel clustering was also developed. The influence was then explored of different partitions of training data sets on the size of the final kernel matrix, on the conditions causing a given cumulative contribution rate, and on the number of subsets. Experimental results showed that a smaller kernel matrix can be obtained when kernel clustering method are used to partition the training dataset. The proposed algorithm can be helpful to reduce the time complexity of the eigen decomposition of a kernel matrix and to improve the speed of feature extraction for test samples.

参考文献/References:

[1]谢永华,陈伏兵,张生亮,杨静宇. 融合小波变换与KPCA的分块人脸特征抽取与识别算法[J]. 中国图象图形学报, 2007, 12(4): 666671.
 XIE Yonghua, CHEN Fubing, ZHANG Shengliang,YANG Jingyu. Features extraction and recognition of intersected human face based on wavelet transform and KPCA[J]. Journal of Image and Graphics, 2007, 12(4): 666671.
 [2]曾庆虎,邱静,刘冠军,谭晓栋. 基于KPCAHSMM设备退化状态识别与故障预测方法研究[J]. 仪器仪表学报, 2009, 30(7): 13411346.
ZENG Qinghu, QIU Jing, LIU Guanjun, TAN Xiaodong. Research on equipment degradation state recognition and fault prognostics method based on KPCAhidden semiMarkov model[J]. Chinese Journal of Scientific Instrument, 2009, 30(7): 13411346.
[3]LI Ying, LEI Xiaogang, BAI Bendu, ZHANG Yanning. Information compression and speckle reduction for multifrequency polarimetric SAR images based on kernel PCA[J]. Journal of Systems Engineering and Electronics, 2008, 19(3): 493498.
 [4]KIM K I, PARK S H, KIM H J. Kernel principal component analysis for texture classification[J]. IEEE Signal Processing Letters, 2001, 8(2): 3941. 
[5]ROSIPAL R, GIROLAMI M, TREJO L J, CICHOCKI A. Kernel PCA for feature extraction and denoising in nonlinear regression[J]. Neural Computing & Applications, 2001, 10(3): 231243.
[6]ZHENG Wenming, ZOU Cairong, ZHAO Li. An improved algorithm for kernel principal component analysis[J]. Neural Processing Letters, 2005, 22(1): 4956.
[7]孔锐,张国宣,施泽生,等. 基于核的K均值聚类[J]. 计算机工程, 2004, 30(11): 1214.
KONG Rui, ZHANG Guoxuan, SHI Zesheng, et al. Kernelbased Kmeans clustering[J]. Computer Engineering, 2004, 30(11): 1214.
[8]The University of Waikaito. WEKA[EB/OL]. [20091202].http://www.cs.waikato.ac.nz/ml/weka.
[9]UCI. Semeion handwritten digit data set[EB/OL]. [20091202]. http://archive.ics.uci.edu/ml/datasets/Semeion+Handwritten+Digit.

相似文献/References:

[1]方鹏,李贤,汪增福.运用核聚类和偏最小二乘回归的歌唱声音转换[J].智能系统学报,2016,11(1):55.[doi:10.11992/tis.201506022]
 FANG Peng,LI Xian,WANG Zengfu.Conversion of singing voice based on kernel clustering and partial least squares regression[J].CAAI Transactions on Intelligent Systems,2016,11(03):55.[doi:10.11992/tis.201506022]

备注/Memo

备注/Memo:
收稿日期:2009-12-19.
基金项目:国家“863”计划资助项目(2007AA04Z423, 2006AA01Z106);
国家自然科学基金资助项目(60576033);
福建省自然科学基金资助项目(2008J04001);
厦门市科技计划资助项目(3502Z20083031).
通信作者:陈 旭.E-mail:shendacx@163.com.
作者简介:
邓 貌,男,1986年生,硕士研究生,主要研究方向为优化算法、神经网络、模式识别等.
 陈 旭,女,1978年生,副研究员,主要研究方向为图像处理、模式识别、神经网络等.
陈天翔,男,1966年生,副研究员,主要研究方向为神经网络、智能计算等
更新日期/Last Update: 2010-07-14