[1]潘 俊,王瑞琴,孔繁胜.基于结构和约束保持的半监督特征选择[J].南京理工大学学报(自然科学版),2014,38(04):518-525.
　Pan Jun,Wang Ruiqin,Kong Fansheng.Semi-supervised feature selection based on structure and constraints preserving[J].Journal of Nanjing University of Science and Technology,2014,38(04):518-525.
点击复制

基于结构和约束保持的半监督特征选择
分享到：

《南京理工大学学报》（自然科学版）[ISSN:1005-9830/CN:32-1397/N]

卷:
38卷
期数:
2014年04期
页码:
518-525
栏目:
出版日期:
2014-08-31

文章信息/Info

Title:
Semi-supervised feature selection based on structure and constraints preserving
作者:
潘 俊1; 王瑞琴2; 孔繁胜3
温州大学 1.信息安全研究所; 2.物理与电子信息工程学院,浙江 温州 325035; 3.浙江大学 计算机科学与技术学院,浙江 杭州 310027
Author(s):
Pan Jun1; Wang Ruiqin2; Kong Fansheng3
1.Institute of Information Security; 2.College of Physics and Electronic Information Engineering, Wenzhou University,Wenzhou 325035,China; 3.College of Computer Science and Technology,Zhejiang University,Hangzhou 310027,China
关键词:
特征选择;  半监督学习;  成对约束;  结构和约束保持;  特征排序;  空间结构;  先验知识
Keywords:
feature selection;  semi-supervised learning;  pairwise constrains;  structure and constraints preserving;  feature ranking;  geometrical structure;  supervision information
分类号:
TP391
摘要:
针对现有特征选择算法较少同时考虑样本的空间结构和先验知识的不足,提出一种基于结构和约束保持的半监督特征选择方法。该方法采用成对约束作为先验知识,同时考虑局部和非局结构,定义了一种新的特征评价准则——结构和约束保持分值。利用大量的无标记样本来学习样本空间结构,利用少量的成对约束信息来学习类内和类间边缘,所选择的特征子集能较好地保持空间结构信息和类属信息。在多个数据集上的实验结果表明,和现有的几种特征排序选择算法相比,所提方法有较好表现。
Abstract:
To overcome the deficiency of most existing feature selection methods which fairly respect both the geometrical structure and the supervision information,a novel approach called semi-supervised feature selection based on structure and constraints preserving is proposed.In this method,both the pairwise constraints and the local and nonlocal structure are taken into account,and a new feature selection criterion,i.e.structure and constraints preserving(SCP)score is defined.The SCP score exploites abundant unlabeled data points to learn the geometrical structure of the data space,and uses a few pairwise constraints to discover the margins of different classes.Those features that can preserve the geometrical structure and pairwise constraints information are selected.Experimental results from several datasets show that the proposed method achieves better performance than the feature ranking selection methods.

参考文献/References:

[1] 杨静宇,金忠,郭跃飞.人脸图像有效鉴别特征抽取与识别[J].南京理工大学学报,2000,24(3):193-198.
Yang Jingyu,Jin Zhong,Guo Yuefei.Efficient discriminant feature extraction and recognition of face images[J].Journal of Nanjing University of Science and Technology,2000,24(3):193-198.
[2]姚旭,王晓丹,张玉玺,等.特征选择方法综述[J].控制与决策,2012,27(2):161-166.
Yao Xu,Wang Xiaodan,Zhang Yuxi,et al.Summary of feature selection algorithms[J].Control and Decision,2012,27(2):161-166.
[3]Li Yun,Lu Baoliang.Feature selection based on loss-margin of nearest neighbor classification[J].Pattern Recognition,2009,42(8):1921-1941.
[4]Zhao Jidong,Lu Ke,He Xiaofei.Locality sensitive semi-supervised feature selection[J].Neurocomputing,2008(71):1842-1849.
[5]蔡哲元,余建国,李先鹏,等.基于核空间距离测度的特征选择[J].模式识别与人工智能,2010,23(2):235-240.
Cai Zheyuan,Yu Jianguo,Li Xianpeng,et al.Feature selection algorithm based on kernel distance measure[J].Pattern Recognition and Artificial Intelligence,2010,23(2):235-240.
[6]业巧林,赵春霞,陈小波.基于正则化技术的对支持向量机特征选择算法[J].计算机研究与发展,2011,48(6):1029-1037.
Ye Qiaolin,Zhao Chunxia,Chen Xiaobo.A feature selection method for twsvm via a regularization technique[J].Journal of Computer Research and Development,2011,48(6):1029-1037.
[7]薛晖,陈松灿.基于局部性正则化推广误差界的特征选择算法[J].模式识别与人工智能,2011,24(4):474-478.
Xue Hui,Chen Songcan.Feature selection based on locality regularized generalization error bound[J].Pattern Recognition and Artificial Intelligence,2011,24(4):474-478.
[8]徐峻岭,周毓明,陈林,等.基于互信息的无监督特征选择[J].计算机研究与发展,2012,49(2):372-382.
Xu Junling,Zhou Yuming,Chen Lin,et al.An unsupervised feature selection approach based on mutual information[J].Journal of Computer Research and Development,2012,49(2):372-382.
[9]Bishop C M.Neural networks for pattern recognition[M].Oxford,UK:Oxford University Press,1995.
[10]He Xiaofei,Cai Deng,Niyoi P.Laplacian score for feature selection[A].Adv Neural Inf Proces Sys[C].Vancouver,Canada:NIPS Foundation,2005:507-514.
[11]Zhang Daoqiang,Chen Songcan,Zhou Zhihua.Constraint score:a new filter method for feature selection with pairwise constraints[J].Pattern Recognition,2008,41(5):1440-1451.
[12]Sun Dan,Zhang Daoqiang.Bagging constraint score for feature selection with pairwise constraints[J].Pattern Recognition,2010,43(6):2106-2118.
[13]He Xiaofei,Yan Shuicheng,Hu Yuxiao,et al.Face recognition using laplacianfaces[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2005,27(3):328-340.
[14]Yang Jian,Zhang Daoqiang,Yang Jingyu,et al.Globally maximizing,locally minimizing:unsupervised discriminant projection with applications to face and palm biometrics[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2007,29(4):650-664.
[15]严云洋,郭志波,陈伏兵,等.融合多尺度多特征的人脸识别方法[J].南京理工大学学报,2009,33(1):48-52.
Yan Yunyang,Guo Zhibo,Chen Fubing,et al.Face recognition based on fusion of multi features with multi scales[J].Journal of Nanjing University of Science and Technology,2009,33(1):48-52.
[16]Chung F R K.Spectral graph theory[M].Fresno,USA:American Mathematical Society,1997.
[17]Blake C L,Merz C J.UCI repository of machine learning databases[EB/OL].[2011-12-26].http://archive.ics.uci.edu/ml.
[18]Samaria F S,Harter A C.Parameterization of a stochastic model for human face identification[A].IEEE Workshop Appl Comput Vision[C].Sarasota,USA:IEEE,Los Alamitos,1994:138-142.
[19]Georghiades A S,Belhumeur P N,Kriegman D J.From few to many:illumination cone models for face recognition under variable lighting and pose[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2001,23(6):643-660.

相似文献/References:

[1]徐敏,张丽萍,朱梧槚.基于Fisher线性判别式的层次文档分类[J].南京理工大学学报(自然科学版),2005,(04):460.
　XU Min~,ZHANG Li-ping~,ZHU Wu-jia~.Hierarchical Document Categorization Based on Fisher Linear Discriminant[J].Journal of Nanjing University of Science and Technology,2005,(04):460.
[2]林 棋,张 宏,李千目.一种基于MA-LSSVM的封装式特征选择算法[J].南京理工大学学报(自然科学版),2016,40(01):10.
　Lin Qi,Zhang Hong,Li Qianmu.Wrapper feature selection algorithm based on MA-LSSVM[J].Journal of Nanjing University of Science and Technology,2016,40(04):10.

备注/Memo

备注/Memo:
收稿日期:2012-12-19 修回日期:2013-02-27
基金项目:浙江省科技计划项目(2012C33086); 浙江省自然科学基金(LQ12F02008)
作者简介:潘俊(1978-),男,博士,讲师,主要研究方向:智能计算,机器学习,数据挖掘,E-mail:panjun@wzu.edu.cn。
引文格式:潘俊,王瑞琴,孔繁胜.基于结构和约束保持的半监督特征选择[J].南京理工大学学报,2014,38(4):518-525.
投稿网址:http://zrxuebao.njust.edu.cn
更新日期/Last Update: 2014-08-31