International Journal of
Physical Sciences

  • Abbreviation: Int. J. Phys. Sci.
  • Language: English
  • ISSN: 1992-1950
  • DOI: 10.5897/IJPS
  • Start Year: 2006
  • Published Articles: 2569

Full Length Research Paper

The algorithm of Fuzzy C-Means clustering based on non-negative matrix factorization

1Key Laboratory Intelligent Computing and Signal Ministry of Education,Anhui University, Hefei, Anhui, China. 2Biomedical Engineering Center, Beijing University of Technology, Beijing, 100022, China.
Email: [email protected]

  •  Accepted: 28 May 2012
  •  Published: 23 November 2012


Clustering analysis is an effective method to discover and identify tumor classes. So, this paper proposes a Fuzzy C-Means clustering (FCM) algorithm based on Non-negative matrix factorization (NMF). Firstly, gene expression profiling (GEP) is simply processed through mean and variance of gene expression, which can then be mapped into a low dimensional space by NMF method. Finally, for discovering and identifying cancer classes, the FCM algorithm is adopted to cluster the GEP. Experimental results show that the NMF reduction dimension method has the capability to resist noise. Compared with Principal component analysis (PCA) method, the NMF reduction dimension method also shows certain advantage.


Key words: Fuzzy C-Means clustering, gene expression profiling, cancer, non-negative matrix factorization.