Title :
Local invariant projection
Author :
Tan, Lu ; Guo, Hongfeng
Author_Institution :
Inst. of Stat. & Math., Shandong Univ. of Finance, Jinan, China
Abstract :
The paper proposed the method of the local invariant projection for dealing with high-dimensional data sets. The method not only has the nature to maintain the geometry and topology structure of the data sets unchanged in the dimension reduction of the high-dimensional data, but the advantages of convenient and rapid calculation in the linear dimension reduction method. What´s more, the regular treatment made the method has good robustness. The results show that the method has the ability to find out the non-linear structure of the data sets.
Keywords :
invariance; signal processing; high dimensional data set; linear dimension reduction method; local invariant projection; topology structure; Eigenvalues and eigenfunctions; Laplace equations; Manifolds; Mathematical model; Nearest neighbor searches; Noise; Principal component analysis; Dimension Reduction; Geometrical Structure; Regularization; Robustness; Topological Structure;
Conference_Titel :
Image and Signal Processing (CISP), 2010 3rd International Congress on
Conference_Location :
Yantai
Print_ISBN :
978-1-4244-6513-2
DOI :
10.1109/CISP.2010.5647620