Abstract

Abstract — In this paper, we briefly review recent advances in blind source separation (BSS) for nonlinear mixing models. After a general introduction to the nonlinear BSS and ICA (independent Component Analysis) problems, we discuss in more detail uniqueness issues, presenting some new results. A fundamental difficulty in the nonlinear BSS problem and even more so in the nonlinear ICA problem is that they are nonunique without extra constraints, which are often implemented by using a suitable regularization. Post-nonlinear mixtures are an important special case, where a nonlinearity is applied to linear mixtures. For such mixtures, the ambiguities are essentially the same as for the linear ICA or BSS problems. In the later part of this paper, various separation techniques proposed for post-nonlinear mixtures and general nonlinear mixtures are reviewed. I. THE NONLINEAR ICA AND BSS PROBLEMS Consider Æ samples of the observed data vector Ü, modeled by

Citations

6517	Elements of information theory - Cover, Thomas - 1991
2717	Neural networks, a comprehensive foundation - Haykin - 1999
1036	Independent component analysis. a new concept - Comon - 1994
1023	Independent Component Analysis - Hyvärinen, Karhunen, et al. - 2001
832	An information-maximization approach to blind separation and blind deconvolution - Bell, Sejnowski - 1995
429	A new learning algorithm for blind signal separation - Amari, Cichocki, et al. - 1996
368	Blind separation of sources, part I: An adaptive algorithm based on neuromimetic architecture - Jutten, Herault - 1991
333	Blind beamforming for non-Gaussian signals - Cardoso, Souloumiac - 1993
325	Equivariant adaptative source separation - Cardoso, Labeld - 1996
316	A fast fixed-point algorithm for independent component analysis - Hyvärinen, Oja - 1997
270	signal separation: statistical principles - Cardoso, “Blind - 1998
235	GTM: the generative topographic mapping - Bishop, Svensen, et al. - 1998
235	I.: Kernel independent component analysis - BACH, JORDAN
227	Lectures on functional equations and their applications - Aczél - 1966
178	Independent factor analysis - Attias - 1999
99	Source separation in post-nonlinear mixtures - Taleb, Jutten - 1999
80	A neural net for blind separation of nonstationary signals - Matsuoka, Ohya, et al. - 1995
79	A blind source separation technique based on second order statistics - Belouchrani, Meraim, et al. - 1997
79	J.F.: Blind separation of instantaneous mixtures of non stationary sources - Pham, Cardoso
77	An unsupervised ensemble learning method for nonlinear dynamic state-space models - Valpola, Karhunen
71	Nonlinear independent component analysis: existence and uniqueness results - Hyvärinen, Pajunen - 1999
71	An Information–Theoretic Approach to Neural Computing - Deco, Obradovic - 1996
70	Robust learning algorithm for blind separation of signals - Cichocki, Unbehauen, et al. - 1994
53	Bayesian nonlinear independent component analysis by multi-layer perceptrons - Lappalainen, Honkela - 2000
52	Blind separation of sources: a non-linear neural algorithm - Burel - 1992
51	Ensemble Learning - Lappalainen, Miskin - 2000
49	Independent Component Analysis: Principles and Practice - Roberts, Everson - 2001
42	Ensemble learning for independent component analysis - Lappalainen - 1999
37	Non-linear blind source separation by self-organizing maps - Pajunen, Hyvdrinen, et al. - 1996
36	Non-linear higher-order statistical decorrelation by volume-conserving neural architectures, Neural Networks - Deco, Brauer - 1995
34	Statistical independence and novelty detection with information preserving nonlinear maps - Parra, Deco, et al. - 1996
34	Information-theoretic approach to blind separation of sources in nonlinear mixture - Yang, Amari, et al. - 1998
29	Building blocks for hierarchical latent variable models - Valpola, Raiko, et al. - 2001
29	ICA mixture models for unsupervised classification of non-gaussian classes and automatic context switching in blind signal separation - Lee, Lewicki, et al. - 2000
28	Blind source separation of nonlinear mixing models - Lee, Koehler, et al. - 1997
28	Hierarchical models of variance sources - Valpola, Harva, et al.
25	Nonlinear independent factor analysis by hierarchical models - Valpola, Östman, et al. - 2003
24	Amuse: a new blind identification algorithm - Tong, Soon, et al. - 1990
24	A maximum likelihood approach to nonlinear blind separation - Pajunen, Karhunen - 1997
22	Temporal BYY learning for state space approach, hidden Markov model and blind source separation - Xu
19	Feature extraction through lococode - Hochreiter, Schmidhuber - 1999
19	Separation of nonlinear mixtures using pattern repulsion - Marques, Almeida - 1999
18	Replicator neural networks for universal optimal source coding - Hecht-Nielsen - 1995
17	On the effect of the form of the posterior approximation in variational learning of ica models - Ilin, Valpola - 2005
17	Variational learning of clusters of undercomplete nonsymmetric independent components - Chan, Lee, et al.
16	and nonlinear ICA based on mutual information - Almeida, “Linear - 2004
16	Nonlinear independent component analysis using ensemble learning theory - Valpola - 2000
16	Nonlinear blind source separation using a radial basis function network - Tan, Wang, et al. - 1994
15	A generic framework for blind source separation in structured nonlinear models - Taleb
15	Faithful representation of separable distributions - Lin, Grier, et al. - 1997