期刊
NEUROCOMPUTING
卷 73, 期 4-6, 页码 968-974出版社
ELSEVIER SCIENCE BV
DOI: 10.1016/j.neucom.2009.08.020
关键词
Graph embedding; Locality preserving projection; L1-norm; Outlier; Dimensionality reduction
资金
- State Key Lab of CAD CG [A0902]
- Zhejiang University
- National Natural Science Foundation of China [60605005, 60975001]
Graph embedding is a general framework for subspace learning. However, because of the well-known outlier-sensitiveness disadvantage of the L2-norm, conventional graph embedding is not robust to outliers which occur in many practical applications. In this paper, an improved graph embedding algorithm (termed LPP-L1) is proposed by replacing L2-norm with L1-norm. In addition to its robustness property, LPP-L1 avoids small sample size problem. Experimental results on both synthetic and real-world data demonstrate these advantages. (C) 2009 Elsevier B.V. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据