A Spectrum Clustering Algorithm Based on Weighted Fuzzy Similar Matrix

Article Preview

Abstract:

Unlike those traditional clustering algorithms, the spectral clustering algorithm can be applied to non-convex sphere of sample spaces and be converged to global optimal. As a entry point that the similar of spectral clustering, introduce improved weighted fuzzy similar matrix to spectral in this paper which avoids influence from parameters changes of fuzzy similar matrix in traditional spectral clustering on clustering effect and improves the effectiveness of clustering. It is more actual and scientific, which is tested based on UCI data set.

You might also be interested in these eBooks

Info:

Periodical:

Advanced Materials Research (Volumes 482-484)

Pages:

2109-2113

Citation:

Online since:

February 2012

Authors:

Export:

Price:

Permissions CCC:

Permissions PLS:

Сopyright:

© 2012 Trans Tech Publications Ltd. All Rights Reserved

Share:

Citation:

[1] ShiJ,MalikJ.Normalized cuts and image segmentation.IEEE Transactionson Pattern Analysis and Machine Intelligence,2000,22(8):888—905.

Google Scholar

[2] Xiaoyan Cai, Guanzhong Dai, Libin Yang. Summary of spectral clustering algorithm[J]. Computer Science, 2008, (7) :14-18.

Google Scholar

[3] Chan P K,Shlag D F,Zien J.Spectral k-way ratio-cut partitioning and clustering [J].IEEE Tran on Computer Aided Design of Integrated Circuits and Systems, 1994,13(9):1088—1096.

DOI: 10.1109/43.310898

Google Scholar

[4] Lihua Liu. The fuzzy clustering method based on weighted similarity[J]. Yunnan Normal University, 2008, (05) :26-30. (In Chinese).

Google Scholar

[5] Xinzhou Wang, Haichi Shu. The construction of the fuzzy similar matrix [J]. Jishou University, 2003, (03):37-41. (In Chinese).

Google Scholar