DOI QR코드

DOI QR Code

An Improved Multiplicative Updating Algorithm for Nonnegative Independent Component Analysis

  • Li, Hui (Institute of Communication Engineering, PLA University of Science and Technology) ;
  • Shen, Yue-Hong (Institute of Communication Engineering, PLA University of Science and Technology) ;
  • Wang, Jian-Gong (Institute of Communication Engineering, PLA University of Science and Technology)
  • Received : 2012.04.12
  • Accepted : 2012.09.25
  • Published : 2013.04.01

Abstract

This paper addresses nonnegative independent component analysis (NICA), with the aim to realize the blind separation of nonnegative well-grounded independent source signals, which arises in many practical applications but is hardly ever explored. Recently, Bertrand and Moonen presented a multiplicative NICA (M-NICA) algorithm using multiplicative update and subspace projection. Based on the principle of the mutual correlation minimization, we propose another novel cost function to evaluate the diagonalization level of the correlation matrix, and apply the multiplicative exponentiated gradient (EG) descent update to it to maintain nonnegativity. An efficient approach referred to as the EG-NICA algorithm is derived and its validity is confirmed by numerous simulations conducted on different types of source signals. Results show that the separation performance of the proposed EG-NICA algorithm is superior to that of the previous M-NICA algorithm, with a better unmixing accuracy. In addition, its convergence speed is adjustable by an appropriate user-defined learning rate.

References

  1. P. Comon and C. Jutten, Handbook of Blind Source Separation: Independent Component Analysis and Applications, Oxford: Academic Press, 2010.
  2. P. Comon, "Independent Component Analysis, A New Concept?" Signal Process., vol. 36, no. 3, Apr. 1994, pp. 287-314. https://doi.org/10.1016/0165-1684(94)90029-9
  3. J. Nascimento and J. Dias, "Does Independent Component Analysis Play a Role in Unmixing Hyperspectral Data?" IEEE Trans. Geosci. Remote Sens., vol. 43, no. 1, Jan. 2005, pp. 175- 187. https://doi.org/10.1109/TGRS.2004.839806
  4. C.-I. Chang, Hyperspectral Data Exploitation: Theory and Applications, Hoboken, NJ: John Wiley & Sons, Inc., 2007.
  5. V.P. Pauce et al., "Unmixing Spectral Data for Space Objects Using Independent Component Analysis and Nonnegative Matrix Factorization," Proc. ATC, 2004.
  6. N. Dobigeon et al., "Bayesian Separation of Spectral Sources under Non-negativity and Full Additivity Constraints," Signal Process., vol. 89, no. 12, Dec. 2009, pp. 2657-2669. https://doi.org/10.1016/j.sigpro.2009.05.005
  7. Y. Sun et al., "Postprocessing and Sparse Blind Source Separation of Positive and Partially Overlapped Data," Signal Process., vol. 91, no. 8, Aug. 2011, pp. 1838-1851. https://doi.org/10.1016/j.sigpro.2011.02.007
  8. R.-C. Henry, "Multivariate Receptor Models - Current Practice and Future Trends," Chemom. Intell. Lab. Syst., vol. 60, no. 1-2, Jan. 2002, pp. 43-48. https://doi.org/10.1016/S0169-7439(01)00184-8
  9. M.-D. Plumbley et al., "Automatic Music Transcription and Audio Source Separation," Cybern. Syst., vol. 33, no. 6, 2002, pp. 603-627. https://doi.org/10.1080/01969720290040777
  10. S. Abdallah and M. Plumbley, "Polyphonic Transcription by Nonnegative Sparse Coding of Power Spectra," Proc. ISMIR, Barcelona, Spain, 2004, pp. 318-325.
  11. M.-D. Plumbley, "Algorithms for Nonnegative Independent Component Analysis," IEEE Trans. Neural Netw., vol. 4, no. 3, May 2003, pp. 534-543.
  12. D.-D. Lee and H.-S. Seung, "Learning the Parts of Objects by Non-negative Matrix Factorization," Nature, vol. 401, Oct. 1999, pp. 788-791. https://doi.org/10.1038/44565
  13. D.-D. Lee and H.-S. Seung, "Algorithms for Non-negative Matrix Factorization," Proc. NIPS, Cambridge, MA: MIT Press, 2001, pp. 556-562.
  14. G.-F. Harpur, Low Entropy Coding with Unsupervised Neural Networks, doctoral dissertation, University of Cambridge, 1997.
  15. M.-D. Plumbley, "Conditions for Nonnegative Independent Component Analysis," IEEE Signal Process. Lett., vol. 9, no. 6, June 2002, pp. 177-180. https://doi.org/10.1109/LSP.2002.800502
  16. E. Oja and M.-D. Plumbley, "Blind Separation of Positive Sources Using Non-negative PCA," Proc. ICABSS, Nara, Japan, 2003, pp. 11-16.
  17. M.-D. Plumbley and E. Oja, "A 'Nonnegative PCA' Algorithm for Independent Component Analysis," IEEE Trans. Neural Netw., vol. 15, no. 1, Jan. 2004, pp. 66-76. https://doi.org/10.1109/TNN.2003.820672
  18. A. Bertrand and M. Moonen, "Blind Separation of Non-negative Source Signals Using Multiplicative Updates and Subspace Projection," Signal Process., vol. 90, no. 10, Oct. 2010, pp. 2877- 2890. https://doi.org/10.1016/j.sigpro.2010.04.015
  19. Z. Yang and J. Laaksonen, "Multiplicative Updates for Nonnegative Projections," Neurocomputing, vol. 71, no. 1-3, Dec. 2007, pp. 363-373. https://doi.org/10.1016/j.neucom.2006.11.023
  20. A. Cichocki and S. Amari, "Extended SMART Algorithms for Non-negative Matrix Factorization," Proc. AISC, Zakopane, Poland, 2006, pp. 548-562.
  21. H. Lutkepohl, Handbook of Matrices, Chichester: John Wiley & Sons Ltd., 1996.
  22. A. Cichocki et al., ICALAB Toolboxes. http://www.bsp.brain. riken.jp/ICALAB