版权说明 帮助中心
首页 > 成果 > 详情

Relative manifold based semi-supervised dimensionality reduction

SCI-EEICSCD
WOS被引频次:4
认领
导出
Link by 维普学术期刊
反馈
分享
QQ微信 微博
成果类型:
期刊论文
作者:
Cai Xianfa;Wen Guihua;Wei Jia;Yu Zhiwen
通讯作者:
Cai, Xianfa
作者机构:
[Cai Xianfa; Wen Guihua; Yu Zhiwen; Wei Jia] School of Computer Science and Engineering, South China University of Technology, Guangzhou, China
[Cai Xianfa] School of Medical Information Engineering, Guangdong Pharmaceutical University, Guangzhou, China
[Cai Xianfa] Shenzhen Key Laboratory of High Performance Data Mining, Shenzhen, China
通讯机构:
[Cai, Xianfa] S China Univ Technol, Sch Med Informat Engn, Guangzhou 510006, Guangdong, Peoples R China.
语种:
英文
关键词:
cognitive law;relative transformation;relative manifold;local reconstruction;semi-supervised learning
关键词(中文):
半监督学习;歧管;降维;非线性结构;拓扑结构;噪声数据;开发利用;认知规律
期刊:
Frontiers of Computer Science
期刊(中文):
中国计算机科学前沿:英文版
ISSN:
2095-2228
年:
2014
卷:
8
期:
6
页码:
923-932
文献类别:
WOS:Article;EI:Journal article (JA)
所属学科:
ESI学科类别:计算机科学;WOS学科类别:Computer Science, Information Systems;Computer Science, Software Engineering;Computer Science, Theory & Methods
入藏号:
WOS:000345385900003;EI:20144500162846
基金类别:
Acknowledgements The research leading to these results was supported by the National Natural Science Foundation of China (Grants No. 61070090, 61273363, 61003174 and 60973083), the Guangdong Natural Science Funds for Distinguished Young Scholar ($2013050014677), the Fundamental Research Funds for the Central Universities (2014G0007), China Postdoctoral Science Foundation (2013M540655), NSFC-Guangdong Joint Fund (U1035004), and Natural Science Foundation of Guangdong Province, China (10451064101004233 and S2012040008022).
机构署名:
本校为第一机构
院系归属:
医药信息工程学院
摘要:
A well-designed graph plays a fundamental role in graph-based semi-supervised learning; however, the topological structure of a constructed neighborhood is unstable in most current approaches, since they are very sensitive to the high dimensional, sparse and noisy data. This generally leads to dramatic performance degradation. To deal with this issue, we developed a relative manifold based semi-supervised dimensionality reduction (RMSSDR) approach by utilizing the relative manifold to construct a better neighborhood graph with fewer short-circuit edges. Based on the relative cognitive law and manifold distance, a relative transformation is used to construct the relative space and the relative manifold. A relative transformation can improve the ability to distinguish between data points and reduce the impact of noise such that it may be more intuitive, and the relative manifold can more truly reflect the manifold structure since data sets commonly exist in a nonlinear structure. Specifically, RMSSDR makes full use of pairwise constraints that can define the edge weights of the neighborhood graph by minimizing the local reconstruction error and can preserve the global and local geometric structures of the data set. The experimental results on face data sets demonstrate that RMSSDR is better than the current state of the art comparing methods in both performance of classification and robustness.
参考文献:
Duda R O, Hart P E, Stork D G. Pattern Classification. New York: John Wiley & Sons, 2001, 566–581
Jolliffe I T. Principal Component Analysis (Springer Series in Statistics). 2nd ed. Springer, 2002
Martinez A M, Kak A C. PCA versus LDA. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2001, 23(2): 228–233
Roweis S, Saul L. Nonlinear dimensionality reduction by locally linear embedding. Science, 2000, 290: 2323–2326
Belkin M, Niyogi P. Laplacian eigenmaps for dimensionality reduction and data representation. Neural Computation, 2003, 15(6): 1373–1396

反馈

验证码:
看不清楚,换一个
确定
取消

成果认领

标题:
用户 作者 通讯作者
请选择
请选择
确定
取消

提示

该栏目需要登录且有访问权限才可以访问

如果您有访问权限,请直接 登录访问

如果您没有访问权限,请联系管理员申请开通

管理员联系邮箱:yun@hnwdkj.com