Abstract
The one-bit-matching conjecture for independent component analysis (ICA) is basically stated as “all the sources can be separated as long as there is one-to-one same-sign-correspondence between the kurtosis signs of all source probability density functions (pdf’s) and the kurtosis signs of all model pdf’s”, which has been widely believed in the ICA community, but not proved completely. Recently, it has been proved that under the assumption of zero skewness for the model pdf’s, the global maximum of a cost function on the ICA problem with the one-bit-matching condition corresponds to a feasible solution of the ICA problem. In this paper, we further study the one-bit-matching conjecture along this direction and prove that all the possible local maximums of this cost function correspond to the feasible solutions of the ICA problem in the case of two sources under the same assumption. That is, the one-bit-matching condition is sufficient for solving the two-source ICA problem via any local ascent algorithm of the cost function.
This work was supported by the Natural Science Foundation of China for Project 60071004 and also by the Hong Kong RGC Earmarked Grant CUHK 4184/03E.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Tong, L., Inouye, Y., Liu, R.: Waveform-preserving blind estimation of multiple independent sources. IEEE Trans. on Signal Processing 41(7), 2461–2470 (1993)
Comon, P.: Independent component analysis–a new concept? Signal Processing 36, 287–314 (1994)
Bell, A., Sejnowski, T.: An information-maximization approach to blind separation and blind deconvolution. Neural Computation 7, 1129–1159 (1995)
Amari, S.I., Cichocki, A., Yang, H.: A new learning algorithm for blind separation of sources. Advances in Neural Information Processing 8, 757–763 (1996)
Oja, E.: ICA learning rules: stationarity,stability, and sigmoids. In: Fyfe, C. (ed.) Proc. of Int. ICSC Workshop on Independence and Artificial Neural Networks (I&ANN 1998), Tenerife, Spain, Febraury 9-10, pp. 97–103. ICSC Acadenic Press, New York (1998)
Cardoso, J.F.: Infomax and maximum likelihood for source separation. IEEE Signal Processing Letters 4, 112–114 (1999)
Xu, L., Cheung, C.C., Amari, S.I.: Learned parametric mixture based ica algorithm. Neurocomputing 22, 69–80 (1998)
Xu, L., Cheung, C.C., Amari, S.I.: Further results on nonlinearity and separation capability of a linear mixture ICA method and learned LPM. In: Fyfe, C. (ed.) Proceedings of the I&ANN 1998, pp. 39–45 (1998)
Girolami, M.: An alternative perspective on adaptive independent component analysis algorithms. Neural Computation 10, 2103–2114 (1998)
Everson, R., Roberts, S.: Independent component analysis: A flexible nonlinearity and decorrelating manifold approach. Neural Computation 11, 1957–1983 (1999)
Lee, T.W., Girolami, M., Sejnowski, T.J.: Independent component analysis using an extended infomax algorithm for mixed subgaussian and supergaussian sources. Neural Computation 11, 417–441 (1999)
Welling, M., Weber, M.: A constrained EM algorithm for independent component analysis. Neural Computation 13, 677–689 (2001)
Cheung, C.C., Xu, L.: Some global and local convergence analysis on the information-theoretic independent component analysis approach. Neurocomputing 30, 79–102 (2000)
Liu, Z.Y., Chiu, K.C., Xu, L.: One-bit-matching conjecture for independent component analysis. Neural Computation 16, 383–399 (2004)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2004 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Ma, J., Liu, Z., Xu, L. (2004). Local Separation Property of the Two-Source ICA Problem with the One-Bit-Matching Condition. In: Yang, Z.R., Yin, H., Everson, R.M. (eds) Intelligent Data Engineering and Automated Learning – IDEAL 2004. IDEAL 2004. Lecture Notes in Computer Science, vol 3177. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-28651-6_101
Download citation
DOI: https://doi.org/10.1007/978-3-540-28651-6_101
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-22881-3
Online ISBN: 978-3-540-28651-6
eBook Packages: Springer Book Archive