An Algebraic Method for Approximate Rank One Factorization of Rank Deficient Matrices

  • Franz J. Király
  • Andreas Ziehe
  • Klaus-Robert Müller
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7191)


In this paper we consider the problem of finding approximate common rank one factors for a set of matrices. Instead of jointly diagonalizing the matrices, we perform calculations directly in the problem intrinsic domain: we present an algorithm, AROFAC, which searches the approximate linear span of the matrices using an indicator function for the rank one factors, finding specific single sources. We evaluate the feasibility of this approach by discussing simulations on generated data and a neurophysiological dataset. Note however that our contribution is intended to be mainly conceptual in nature.


Loss Function Singular Vector Blind Source Separation Constant Modulus Algorithm Joint Diagonalization 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Bießmann, F., Meinecke, F.C., Gretton, A., Rauch, A., Rainer, G., Logothetis, N.K., Müller, K.R.: Temporal kernel CCA and its application in multimodal neuronal data analysis. Machine Learning 79(1-2), 5–27 (2010)MathSciNetCrossRefGoogle Scholar
  2. 2.
    von Bünau, P., Meinecke, F.C., Király, F.J., Müller, K.R.: Finding stationary subspaces in multivariate time series. Phys. Rev. Lett. 103(21), 214101 (2009)CrossRefGoogle Scholar
  3. 3.
    Cardoso, J.F., Souloumiac, A.: Blind beamforming for non Gaussian signals. IEE - Proceedings -F 140(6), 362–370 (1993)Google Scholar
  4. 4.
    Cardoso, J.F., Souloumiac, A.: Jacobi angles for simultaneous diagonalization. SIAM Journal on Matrix Analysis and Applications 17(1), 161–164 (1996)MathSciNetCrossRefzbMATHGoogle Scholar
  5. 5.
    van Der Veen, A.J., Paulraj, A.: An analytical constant modulus algorithm. IEEE Trans. Signal Processing 44(5), 1–19 (1996)CrossRefGoogle Scholar
  6. 6.
    Lathauwer, L.D.: A link between the canonical decomposition in multilinear algebra and simultaneous matrix diagonalization. SIAM J. Matrix Analysis Applications 28(3), 642–666 (2006)MathSciNetCrossRefzbMATHGoogle Scholar
  7. 7.
    Müller, K.R., Tangermann, M., Dornhege, G., Krauledat, M., Curio, G., Blankertz, B.: Machine learning for real-time single-trial analysis: From brain-computer interfacing to mental state monitoring. Journal of Neuroscience Methods 167, 82–90 (2008)CrossRefGoogle Scholar
  8. 8.
    Nocedal, J.: Updating quasi-Newton matrices with limited storage. Mathematics of Computation 35(151), 773–782 (1980)MathSciNetCrossRefzbMATHGoogle Scholar
  9. 9.
    Nolte, G., Bai, O., Wheaton, L., Mari, Z., Vorbach, S., Hallett, M.: Identifying true brain interaction from EEG data using the imaginary part of coherency. Clinical Neurophysiology 115(10), 2292–2307 (2004), CrossRefGoogle Scholar
  10. 10.
    Nolte, G., Meinecke, F.C., Ziehe, A., Müller, K.R.: Identifying interactions in mixed and noisy complex systems. Phys. Rev. E 73, 051913 (2006), CrossRefGoogle Scholar
  11. 11.
    Nolte, G., Ziehe, A., Meinecke, F., Müller, K.-R.: Analyzing coupled brain sources: Distinguishing true from spurious interaction. In: Weiss, Y., Schölkopf, B., Platt, J. (eds.) Advances in Neural Information Processing Systems, vol. 18, pp. 1027–1034. MIT Press, Cambridge (2006)Google Scholar
  12. 12.
    Pham, D.T.: Joint approximate diagonalization of positive definite matrices. SIAM J. on Matrix Anal. and Appl. 22, 1136–1152 (2001)MathSciNetCrossRefzbMATHGoogle Scholar
  13. 13.
    Schölkopf, B., Smola, A.J., Müller, K.R.: Nonlinear component analysis as a kernel eigenvalue problem. Neural Computation 10(5), 1299–1319 (1998)CrossRefGoogle Scholar
  14. 14.
    van der Veen, A.: Joint diagonalization via subspace fitting techniques. In: Proc. ICASSP, vol. 5 (2001)Google Scholar
  15. 15.
    Yeredor, A.: Non-orthogonal joint diagonalization in the least-squares sense with application in blind source separation. IEEE Trans. Signal Processing 50(7), 1545–1553 (2002)MathSciNetCrossRefGoogle Scholar
  16. 16.
    Yeredor, A.: On using exact joint diagonalization for noniterative approximate joint diagonalization. IEEE Signal Processing Letters 12(9), 645–648 (2005)CrossRefGoogle Scholar
  17. 17.
    Ziehe, A., Laskov, P., Nolte, G., Müller, K.R.: A fast algorithm for joint diagonalization with non-orthogonal transformations and its application to blind source separation. Journal of Machine Learning Research 5, 777–800 (2004)MathSciNetzbMATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Franz J. Király
    • 1
  • Andreas Ziehe
    • 1
  • Klaus-Robert Müller
    • 1
  1. 1.Machine Learning GroupTechnische Universität BerlinBerlinGermany

Personalised recommendations