Title :
RPCL-based local PCA algorithm
Author :
Liu, Zhiyong ; Xu, Lei
Author_Institution :
Dept. of Comput. Sci. & Eng., Chinese Univ. of Hong Kong, China
Abstract :
Mining local structure is important in data analysis. Gaussian mixture is able to describe local structure through covariance matrices, but when used on high-dimensional data, specifying such a large number of d(d+1)/2 free elements in each covariance matrix is difficult. By constraining the covariance matrix in decomposed orthonormal form, we propose a Local PCA algorithm to tackle this problem with the help of RPCL (Rival Penalized Competitive Learning), which can automatically determine the number of local structures
Keywords :
Gaussian processes; covariance matrices; data mining; unsupervised learning; Gaussian mixture; RPCL competitive learning; RPCL-based Local PCA Algorithm; Rival Penalized Competitive Learning; covariance matrices; data analysis; decomposed orthonormal form; high dimensional data; local structure; local structure mining; local structures; Clustering algorithms; Computer science; Costs; Covariance matrix; Data analysis; Principal component analysis; Shape;
Conference_Titel :
Data Mining, 2001. ICDM 2001, Proceedings IEEE International Conference on
Conference_Location :
San Jose, CA
Print_ISBN :
0-7695-1119-8
DOI :
10.1109/ICDM.2001.989582
