关键词: 主成分分析, 局部切空间, 流形学习

Abstract: Linear local tangent space alignment (LLTSA) algorithm is a non-linear dimension reducing method which can easily apply to recognition problems, it pays attention on the local geometric structure of data, but it neglects the global information of data. In this paper an improved LLTSA algorithm which based on principal component analysis (PCA) is proposed, this method take the global structure of sample into consider and contain a better reducing dimension result. In the classical experiment of 3D manifold and MNSIT image dataset script recognize, PLLTSA has a higher recognition rate contrast to PCA, LPP and LLTSA, which verify the effectiveness of PLLTSA.

Key words: Principal component analysis, Local tangent space, Manifold learning