A generalized least-squares approach regularized with graph embedding for dimensionality reduction
文献类型:期刊论文
作者 | Shen, Xiang-Jun2; Liu, Si-Xing2; Bao, Bing-Kun3; Pan, Chun-Hong4; Zha, Zheng-Jun5; Fan, Jianping1 |
刊名 | PATTERN RECOGNITION |
出版日期 | 2020-02-01 |
卷号 | 98页码:10 |
ISSN号 | 0031-3203 |
关键词 | Dimensionality reduction Graph embedding Subspace learning Least-squares |
DOI | 10.1016/j.patcog.2019.107023 |
通讯作者 | Shen, Xiang-Jun(xjshen@ujs.edu.cn) ; Zha, Zheng-Jun(zhazj@ustc.edu.cn) |
英文摘要 | In current graph embedding methods, low dimensional projections are obtained by preserving either global geometrical structure of data or local geometrical structure of data. In this paper, the PCA (Principal Component Analysis) idea of minimizing least-squares reconstruction errors is regularized with graph embedding, to unify various local manifold embedding methods within a generalized framework to keep global and local low dimensional subspace. Different from the well-known PCA method, our proposed generalized least-squares approach considers data distributions together with an instance penalty in each data point. In this way, PCA is viewed as a special instance of our proposed generalized least squares framework for preserving global projections. Applying a regulation of graph embedding, we can obtain projection that preserves both intrinsic geometrical structure and global structure of data. From the experimental results on a variety of face and handwritten digit recognition, our proposed method has advantage of superior performances in keeping lower dimensional subspaces and higher classification results than state-of-the-art graph embedding methods. (C) 2019 Elsevier Ltd. All rights reserved. |
WOS关键词 | PRESERVING PROJECTIONS ; EIGENMAPS ; FRAMEWORK |
资助项目 | National Natural Science Foundation of China[61572240] ; National Natural Science Foundation of China[61622211] ; National Natural Science Foundation of China[61620106009] ; National Natural Science Foundation of China[61572503] ; National Natural Science Foundation of China[61872424] ; National Natural Science Foundation of China[6193000388] ; Open Project Program of the National Laboratory of Pattern Recognition (NLPR)[201600005] |
WOS研究方向 | Computer Science ; Engineering |
语种 | 英语 |
出版者 | ELSEVIER SCI LTD |
WOS记录号 | WOS:000497600300024 |
资助机构 | National Natural Science Foundation of China ; Open Project Program of the National Laboratory of Pattern Recognition (NLPR) |
源URL | [http://ir.ia.ac.cn/handle/173211/29385] |
专题 | 自动化研究所_模式识别国家重点实验室_多媒体计算与图形学团队 |
通讯作者 | Shen, Xiang-Jun; Zha, Zheng-Jun |
作者单位 | 1.Univ N Carolina, Dept Comp Sci, Charlotte, NC 28223 USA 2.JiangSu Univ, Sch Comp Sci & Commun Engn, Nanjing 212013, Jiangsu, Peoples R China 3.Nanjing Univ Posts & Telecommun, Nanjing, Jiangsu, Peoples R China 4.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 5.Univ Sci & Technol China, Sch Informat Sci & Technol, Hefei, Anhui, Peoples R China |
推荐引用方式 GB/T 7714 | Shen, Xiang-Jun,Liu, Si-Xing,Bao, Bing-Kun,et al. A generalized least-squares approach regularized with graph embedding for dimensionality reduction[J]. PATTERN RECOGNITION,2020,98:10. |
APA | Shen, Xiang-Jun,Liu, Si-Xing,Bao, Bing-Kun,Pan, Chun-Hong,Zha, Zheng-Jun,&Fan, Jianping.(2020).A generalized least-squares approach regularized with graph embedding for dimensionality reduction.PATTERN RECOGNITION,98,10. |
MLA | Shen, Xiang-Jun,et al."A generalized least-squares approach regularized with graph embedding for dimensionality reduction".PATTERN RECOGNITION 98(2020):10. |
入库方式: OAI收割
来源:自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。