Regularised nonlinear blind signal separation using sparsely connected network
IEE Proceedings - Vision Image and Signal ProcessingPublished 1 January 2005
Wai Lok Woo, S.S. Dlay
Citations17
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
A nonlinear approach based on the Tikhonov regularised cost function is presented for blind signal separation of nonlinear mixtures and it is shown that this approach can be jointly used to extract independent components while constraining the overall perceptron network to be as sparse as possible.
Abstract
A nonlinear approach based on the Tikhonov regularised cost function is presented for blind signal separation of nonlinear mixtures. The proposed approach uses a multilayer perceptron as the nonlinear demixer and combines both information theoretic learning and structural complexity learning into a single framework. It is shown that this approach can be jointly used to extract independent components while constraining the overall perceptron network to be as sparse as possible. The update algorithm for the nonlinear demixer is subsequently derived using the new cost function. Sparseness in the network connection is utilised to determine the total number of layers required in the multilayer perceptron and to prevent the nonlinear demixer from outputting arbitrary independent components. Experiments are meticulously conducted to study the performance of the new approach and the outcomes of these studies are critically assessed for performance comparison with existing methods.
Keywords
ChemistryComputer Science
NatureLearning representations by back-propagating errors
30,885 Citations1986David E. Rumelhart, Geoffrey E. Hinton +1 more
Back-propagation repeatedly adjusts the weights of the connections in the network so as to minimize a measure of the difference between the actual output vector of the net and the desired output vector, which helps to represent important features of the task domain.
Mathematics of Control Signals and SystemsApproximation by superpositions of a sigmoidal function
13,684 Citations1989George Cybenko
John Wiley & Sons, Inc. eBooksAdaptive Blind Signal and Image Processing: Learning Algorithms and Applications
1,286 Citations2002Andrzej Cichocki, Шун-ичи Амари
This volume unifies and extends the theories of adaptive blind signal and image processing and provides practical and efficient algorithms for blind source separation, Independent, Principal, Minor Component Analysis, and Multichannel Blind Deconvolution (MBD) and Equalization.
Neural NetworksNonlinear independent component analysis: Existence and uniqueness results
526 Citations1999Aapo Hyvärinen, P. Pajunen
It is shown that if the space of mixing functions is not limited there exists always an infinity of solutions, and that for two dimensions, the solution is unique up to a rotation, if the mixing function is constrained to be a conformal mapping together with some other assumptions.
IEEE Transactions on Signal ProcessingSource separation in post-nonlinear mixtures
412 Citations1999Anas Abu Taleb, Christian Jutten
It is proved that in the general case, it is not possible to separate the sources without nonlinear distortion, and this work focuses on specific nonlinear mixtures known as post-nonlinear mixture, which are realistic models in many situations and emphasize interesting properties i.e., in such nonlinearmixtures, sources can be estimated with the same indeterminacies as in instantaneous linear mixtures.
Neural NetworksBlind separation of sources: A nonlinear neural algorithm
215 Citations1992Gilles Burel
A new algorithm, based on ideas of back propagation learning, is proposed for source separation, and the algorithm can deal even with nonlinear mixtures.
IEEE Transactions on Neural NetworksFeedback stabilization using two-hidden-layer nets
204 Citations1992Eduardo D. Sontag
It is shown that, contrary to what might have been expected from the well-known representation theorems, three-layer nets are not sufficient for stabilization, but four-layernets are enough, assuming that threshold processors are used.
IEEE Transactions on Neural NetworksNonlinear blind source separation using a radial basis function network
162 Citations2001Ying Tan, Jun Wang +1 more
By virtue of the RBF neural network, this proposed approach takes advantage of high learning convergence rate of weights in the hidden layer and output layer, natural unsupervised learning characteristics, modular structure, and universal approximation capability.
Signal ProcessingInformation–theoretic approach to blind separation of sources in non-linear mixture
153 Citations1998Howard H. Yang, Шун-ичи Амари +1 more
A two-layer perceptron is used as a de-mixing system to separate sources in non-linear mixture to derive the learning equations for the hidden layer.
Advances in Nonlinear Blind Source Separation
123 Citations2003Christian Jutten, Juha Karhunen
In this paper, recent advances in blind source separation (BSS) for nonlinear mixing models are reviewed and various separation techniques proposed for post-nonlinear mixtures and general non linear mixtures are reviewed.
Blind source separation of nonlinear mixing models
95 Citations2002Te-Won Lee, Bert-Uwe Koehler +1 more
A new set of learning rules for the nonlinear blind source separation problem based on the information maximization criterion is presented and a parametric sigmoidal nonlinearity and higher order polynomials are proposed.
IEEE Transactions on Neural NetworksNonlinear blind source separation using kernels
70 Citations2003Dominique Martinez, Austin M. Bray
A new method is derived for solving nonlinear blind source separation problems by exploiting second-order statistics in a kernel induced feature space using the kernel trick originally applied in support vector machines (SVMs).
IEEE Transactions on Signal ProcessingA generic framework for blind source separation in structured nonlinear models
60 Citations2002Anas Abu Taleb
It is shown how a proper use of the relative gradient leads to equivariant adaptive algorithms for both batch and stochastic algorithms in blind source separation in nonlinear models.
Nonlinear Blind Source Separation by Variational Bayesian Learning
36 Citations2003Harri Valpola, Erkki Oja +3 more
This paper first considers a static nonlinear mixing model, with a successful application to real-world speech data compression, and discusses extraction of sources from nonlinear dynamic processes, and detection of abrupt changes in the process dynamics.
Nonlinear blind source separation using kernel feature spaces
29 Citations2001Stefan Harmeling, Kawanabe M Ziehe A +5 more
A kernel-based blind source separation (BSS) algorithm that can perform nonlinear BSS for general invertible nonlinearities and a criterion that allows to identify the original sources is proposed.
IEE Proceedings - Vision Image and Signal ProcessingBlind restoration of nonlinearly mixed signals using multilayer polynomial neural network
27 Citations2004Wai Lok Woo, L.C. Khor
A new mathematical model of the nonlinear mixing system has been developed culminating in the formulation of a stable unique inverse solution, which has an identical structure to the multilayer neural network.
Neural Information Processing SystemsSource Separation and Density Estimation by Faithful Equivariant SOM
27 Citations1996Juan K. Lin, Jack D. Cowan +1 more
BLIND SOURCE SEPARATION IN NONLINEAR MIXTURES BY ADAPTIVE SPLINE NEURAL NETWORKS
12 Citations2001M. Solazzi, Raffaele Parisi +1 more
A novel paradigm for blind source separation in the presence of nonlinear mixtures is presented and described, employing a neural model based on adaptive B-spline functions that achieves signal separation through an information maximization criterion.