Abstract
In this paper a hybrid approach, based on a functional network and a neural network, for post-nonlinear independent component analysis is presented. In order to obtain the independence among the outputs, it was used as cost function a measure based on Renyi’s quadratic entropy and Cauchy-Sc hw artz inequality. Also, the Kernel method was used for nonparametric estimation of the probability density function. The experimental results corroborated the soundness of the approach and a comparative study with a neural network showed its superior performance.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Burel, G.: A non-linear neural algorithm. Meural Networks 5 (1992) 937–947
Castillo, E.: Functional Networks. Neural Processing Letters, 7(3) (1998) 151–159
Castillo, E., Cobo, A., Gutierrez, J. M., Pruneda, R. E.: Functional Networks with Applications. A Neural-Based Paradigm. Kluwer Academic Publishers, Dordrecht (1998)
Comon, P.: Independent Component Analysis-a new concept?. Signal Processing 36 (1994) 287–314
Duda, R. O., Hart, P. E.: Pattern Classification and Scene Analysis. John Wiley & Sons, New York (1973)
Hyvarinen, A., Pajunen, P.: Nonlinear Independent Component Analysis: Existence and Uniqueness results. Neural Networks 12(3) (1999) 429–439
Jutten, C., Herault, J.: Blind separation of sources, part I: An adaptive algorithm based on neuromimetic architecture. Signal Processing 24 (1991) 1–10
Pajunen, P.: Nonlinear independent component analysis by self-organizing maps. Proceedings ICANN (1996) 815–819
Principe, J. C., Fisher III, J., Xu, D.: Information Theoretic Learning. In S. Haykin (Ed.): Unsupervised Adaptive Filtering, Vol. I. John Wiley & Sons, New York (2000)
Principe, J. C., Xu, D.: Information Theoretic Learning using Renyi’s Quadratic Entropy. Proceedings ICA’99 (1999) 407–412
Schobben, D. W. E., Torkkola, K., Smaragdis, P.: Evaluation of Blind Signal Separation Methods. Proceedings ICA’99 (1999) 261–266
Silverman, B. W.: Density estimation for statistics and data analysis. Chapman & Hall, London (1986)
Taleb, A., Jutten, C.: Nonlinear source separation: The post-nonlinear mixtures. Proceedings ESANN (1997) 279–284
Taleb, A., Jutten, C.: Source separation in post-nonlinear mixtures. IEEE Trans. on Signal Processing 47(10) (1999) 2807–2820
Xu, D., Principe, J. C., Fisher III, J., Wu, H.-C.: A Novel Measure for Independent Component Analysis (ICA). Proceedings ICASSP, Vol. II (1998) 1161–1164
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2001 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Romero, O.F., Berdiñas, B.G., Betanzos, A.A. (2001). A Functional-Neural Network for Post-Nonlinear Independent Component Analysis. In: Mira, J., Prieto, A. (eds) Connectionist Models of Neurons, Learning Processes, and Artificial Intelligence. IWANN 2001. Lecture Notes in Computer Science, vol 2084. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-45720-8_34
Download citation
DOI: https://doi.org/10.1007/3-540-45720-8_34
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-42235-8
Online ISBN: 978-3-540-45720-6
eBook Packages: Springer Book Archive