一种正交流形保持投影方法
首发时间:2010-09-13
摘要:特征提取是模式识别领域中一个重要的研究方向,保局投影是一种非常有效的特征提取方法.为了克服保局投影的缺陷,提出了正交流形保持投影方法,在保局投影目标函数中引入了数据的非近邻信息, 有效地保持了数据的局部流形结构和全局流形结构;采用格拉姆-施密特正交化过程获取正交投影基向量,解决了保局投影非正交问题。在ORL和Yale人脸数据库上进行了实验,实验结果验证了该方法的有效性。
关键词: 模式识别与智能系统 保局投影 流形保持投影 正交流形保持投影 格拉姆-施密特正交变换
For information in English, please click here
Orthogonal Manifold Preserving Projections
Abstract:Feature extraction is one of important research areas in pattern recognition, locality preserving projections (LPP) is one of the most effective methods for feature extraction. In order to overcome the shortcomings of LPP, we propose a new feature extraction methods--- orthogonal manifold preserving projections (OMPP). The non-locality information of data are incorporates to the objective function, in contrast to LPP, which preserve the local and global structure of the data manifold. Gram-Schmidt orthogonalization is used to determine basic orthogonal vectors,which solve to non-orthogonal problem of LPP. Experiments on ORL and Yale face databases are performed to test and evaluate the proposed algorithm, and the results have shown that the proposed method is effective.
Keywords: Pattern Recognition and Intelligent Systems locality preserving projections manifold preserving projections orthogonal manifold preserving projections Gram-Schmidt orthogonalization
论文图表:
引用
No.4384685541151128****
同行评议
共计0人参与
勘误表
一种正交流形保持投影方法
评论
全部评论0/1000