Advertisement

Hierarchical ALS Algorithms for Nonnegative Matrix and 3D Tensor Factorization

  • Andrzej Cichocki
  • Rafal Zdunek
  • Shun-ichi Amari
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4666)

Abstract

In the paper we present new Alternating Least Squares (ALS) algorithms for Nonnegative Matrix Factorization (NMF) and their extensions to 3D Nonnegative Tensor Factorization (NTF) that are robust in the presence of noise and have many potential applications, including multi-way Blind Source Separation (BSS), multi-sensory or multi-dimensional data analysis, and nonnegative neural sparse coding. We propose to use local cost functions whose simultaneous or sequential (one by one) minimization leads to a very simple ALS algorithm which works under some sparsity constraints both for an under-determined (a system which has less sensors than sources) and over-determined model. The extensive experimental results confirm the validity and high performance of the developed algorithms, especially with usage of the multi-layer hierarchical NMF. Extension of the proposed algorithm to multidimensional Sparse Component Analysis and Smooth Component Analysis is also proposed.

Keywords

Blind Source Separation Nonnegative Matrix Factorization Alternate Little Square Nonlinear Projection Alternate Little Square Algorithm 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Cichocki, A., Amari, S.: Adaptive Blind Signal And Image Processing (New revised and improved edition). John Wiley, New York (2003)Google Scholar
  2. 2.
    Dhillon, I., Sra, S.: Generalized nonnegative matrix approximations with Bregman divergences. In: Neural Information Proc. Systems, Vancouver, Canada (2005)Google Scholar
  3. 3.
    Hazan, T., Polak, S., Shashua, A.: Sparse image coding using a 3D non-negative tensor factorization. In: International Conference of Computer Vision (ICCV), pp. 50–57 (2005)Google Scholar
  4. 4.
    Heiler, M., Schnoerr, C.: Controlling sparseness in non-negative tensor factorization. In: Leonardis, A., Bischof, H., Pinz, A. (eds.) ECCV 2006. LNCS, vol. 3951, pp. 56–67. Springer, Heidelberg (2006)CrossRefGoogle Scholar
  5. 5.
    Hoyer, P.: Non-negative matrix factorization with sparseness constraints. Journal of Machine Learning Research 5, 1457–1469 (2004)MathSciNetGoogle Scholar
  6. 6.
    Morup, M., Hansen, L.K., Herrmann, C.S., Parnas, J., Arnfred, S.M.: Parallel factor analysis as an exploratory tool for wavelet transformed event-related EEG. NeuroImage 29, 938–947 (2006)CrossRefGoogle Scholar
  7. 7.
    Smilde, A., Bro, R., Geladi, P.: Multi-way Analysis: Applications in the Chemical Sciences. John Wiley and Sons, New York (2004)Google Scholar
  8. 8.
    Oja, E., Plumbley, M.D.: Blind separation of positive sources by globally convergent gradient search. Neural Computation 16, 1811–1825 (2004)zbMATHCrossRefGoogle Scholar
  9. 9.
    Lee, D.D., Seung, H.S.: Learning the parts of objects by nonnegative matrix factorization. Nature 401, 788–791 (1999)CrossRefGoogle Scholar
  10. 10.
    Berry, M., Browne, M., Langville, A., Pauca, P., Plemmons, R.: Algorithms and applications for approximate nonnegative matrix factorization. Computational Statistics and Data Analysis (in press, 2006)Google Scholar
  11. 11.
    Cichocki, A., Amari, S., Zdunek, R., Kompass, R., Hori, G., He, Z.: Extended SMART algorithms for non-negative matrix factorization. In: Rutkowski, L., Tadeusiewicz, R., Zadeh, L.A., Zurada, J.M. (eds.) ICAISC 2006. LNCS (LNAI), vol. 4029, pp. 548–562. Springer, Heidelberg (2006)CrossRefGoogle Scholar
  12. 12.
    Kim, M., Choi, S.: Monaural music source separation: Nonnegativity, sparseness, and shift-invariance. In: Rosca, J., Erdogmus, D., Príncipe, J.C., Haykin, S. (eds.) ICA 2006. LNCS, vol. 3889, pp. 617–624. Springer, Heidelberg (2006)CrossRefGoogle Scholar
  13. 13.
    Zdunek, R., Cichocki, A.: Non-negative matrix factorization with quasi-Newton optimization. In: Rutkowski, L., Tadeusiewicz, R., Zadeh, L.A., Zurada, J.M. (eds.) ICAISC 2006. LNCS (LNAI), vol. 4029, pp. 870–879. Springer, Heidelberg (2006)CrossRefGoogle Scholar
  14. 14.
    Zdunek, R., Cichocki, A.: Nonnegative matrix factorization with constrained second-order optimization. Signal Processing 87, 1904–1916 (2007)CrossRefGoogle Scholar
  15. 15.
    Cichocki, A., Zdunek, R.: Multilayer nonnegative matrix factorization. Electronics Letters 42, 947–948 (2006)CrossRefGoogle Scholar
  16. 16.
    Murray, J.F., Kreutz-Delgado, K.: Learning sparse overcomplete codes for images. Journal of VLSI Signal Processing 45, 97–110 (2006)CrossRefGoogle Scholar
  17. 17.
    Kreutz-Delgado, K., Murray, J.F., Rao, B.D., Engan, K., Lee, T.W., Sejnowski, T.J.: Dictionary learning algorithms for sparse representation. Neural Computation 15, 349–396 (2003)zbMATHCrossRefGoogle Scholar
  18. 18.
    Cichocki, A., Zdunek, R.: NTFLAB for Signal Processing. Technical report, Laboratory for Advanced Brain Signal Processing, BSI, RIKEN, Saitama, Japan (2006)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Authors and Affiliations

  • Andrzej Cichocki
    • 1
  • Rafal Zdunek
    • 2
  • Shun-ichi Amari
    • 3
  1. 1.Dept. of EE, Warsaw University of Technology, and IBS PAN WarsawPoland
  2. 2.Institute of Telecommunications, Teleinformatics and Acoustics, Wroclaw University of TechnologyPoland
  3. 3.RIKEN Brain Science Institute, Wako-shi, SaitamaJapan

Personalised recommendations