[1]王靖.基于离群点检测的鲁棒局部切空间排列方法[J].华侨大学学报(自然科学版),2008,29(4):522-526.[doi:10.11830/ISSN.1000-5013.2008.04.0522]
　WANG Jing.Robust Local Tangent Space Alignment Based on Outlier Detection[J].Journal of Huaqiao University(Natural Science),2008,29(4):522-526.[doi:10.11830/ISSN.1000-5013.2008.04.0522]
点击复制

基于离群点检测的鲁棒局部切空间排列方法()
分享到：

《华侨大学学报（自然科学版）》[ISSN:1000-5013/CN:35-1079/N]

卷:
第29卷
期数:
2008年第4期
页码:
522-526
栏目:
出版日期:
2008-10-20

文章信息/Info

Title:
Robust Local Tangent Space Alignment Based on Outlier Detection
文章编号:
1000-5013(2008)04-0522-05
作者:
王靖
华侨大学信息科学与工程学院
Author(s):
WANG Jing
College of Information Science and Engineering, Huaqiao University, Quanzhou 362021, China
关键词:
鲁棒;  离群点;  流形学习;  局部切空间排列
Keywords:
robust;  outlier;  manifold learning;  local tangent space alignment
分类号:
TP181
DOI:
10.11830/ISSN.1000-5013.2008.04.0522
文献标志码:
A
摘要:
研究局部切空间排列方法(LTSA)对离群点的敏感性,提出一种基于离群点检测的鲁棒局部切空间排列方法(RLTSA).该方法用样本点到切空间的投影距离检测离群点.在构造样本点局部邻域时,RLTSA尽可能排除离群点,以构造稳定的局部邻域,而对离群点,RLTSA把它们投影到更高维的切空间,以减少离群点的投影距离.模拟实验和实际例子说明,新方法能提高局部切空间排列方法处理离群样本点的能力.
Abstract:
The paper focuses on the sensitivity of local tangent space alignment(LTSA) to outliers,and presents a robust local tangent space alignment(RLTSA) based on outlier detection.RLTSA detects the outliers by the projection distances of the sample points onto their tangent spaces.When constructing the neighborhoods of the sample points,RLTSA removes the outliers to construct stably local neighborhood.For the outliers,RLTSA projects them onto higher dimensional tangent spaces to reduce their projection distances.Simulation and real examples show that the new approach can improve the ability of LTSA to deal with outliers.

参考文献/References:

[1] TENENBAUM J, SILVA V D, LANGFORD J. A global geometric framework for nonlinear dimension reduction [J]. Science, 2000.2319-2323.doi:10.1126/science.290.5500.2268.
[2] ROWEIS S, SAUL L. Nonlinear dimensionality reduction by locally linear embedding [J]. Science, 2000.2323-2326.
[3] ZHANG Z, WANG J. MLLE: Modified locally linear embedding using multiple weights [A]. 2007.
[4] BELKIN M, NIYOGI P. Laplacian eigenmaps and spectral techniques for embedding and clustering [M]. Cambridge, ma:the Mit Press, 2001.
[5] ZHANG Z, ZHA H. Principal manifolds and nonlinear dimensionality reduction via tangent space alignment [J]. SIAM Journal on Scientific Computing, 2005(1):313-338.
[6] SILVA V, TENENBAUM J B. Global versus local methods in nonlinear dimensionality reduction [M]. Cambridge, ma:the Mit Press, 2003.705-712.
[7] ZHANG Z, ZHA H. Local linear smoothing for nonlinear manifold learning [R]. PA:Pennsylvania Statue University, 2003.
[8] CHANG H, YEUNG D Y. Robust locally linear embedding [J]. Pattern Recognition, 2006(6):1053-1065.doi:10.1016/j.patcog.2005.07.011.

备注/Memo

备注/Memo:
福建省青年科技人才创新基金(2007F3067); 华侨大学高层次人才科研启动项目(06BS304)
更新日期/Last Update: 2014-03-23