DocumentCode :
2771442
Title :
Non-negative Laplacian Embedding
Author :
Luo, Dijun ; Ding, Chris ; Huang, Heng ; Li, Tao
Author_Institution :
Comput. Sci. & Eng. Dept., Univ. of Texas at Arlington, Arlington, TX, USA
fYear :
2009
fDate :
6-9 Dec. 2009
Firstpage :
337
Lastpage :
346
Abstract :
Laplacian embedding provides a low dimensional representation for a matrix of pairwise similarity data using the eigenvectors of the Laplacian matrix. The true power of Laplacian embedding is that it provides an approximation of the ratio cut clustering. However, ratio cut clustering requires the solution to be nonnegative. In this paper, we propose a new approach, nonnegative Laplacian embedding, which approximates ratio cut clustering in a more direct way than traditional approaches. From the solution of our approach, clustering structures can be read off directly. We also propose an efficient algorithm to optimize the objective function utilized in our approach. Empirical studies on many real world datasets show that our approach leads to more accurate ratio cut solution and improves clustering accuracy at the same time.
Keywords :
approximation theory; eigenvalues and eigenfunctions; matrix decomposition; Laplacian matrix; eigenvector; low dimensional representation; nonnegative Laplacian embedding; pairwise similarity data; ratio cut clustering; Clustering algorithms; Computer science; Data engineering; Data mining; Information retrieval; Laplace equations; Machine learning; Matrix decomposition; Power engineering and energy; Vectors; Clustering; Dimension reduction; Laplacian Embedding; Non-negative Matrix Factorization;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Data Mining, 2009. ICDM '09. Ninth IEEE International Conference on
Conference_Location :
Miami, FL
ISSN :
1550-4786
Print_ISBN :
978-1-4244-5242-2
Electronic_ISBN :
1550-4786
Type :
conf
DOI :
10.1109/ICDM.2009.74
Filename :
5360259
Link To Document :
بازگشت