An extension of the infomax algorithm of Bell and Sejnowski (1995) is presented that is able to blindly separate mixed signals with sub- and super-Gaussian source distributions. This was achieved by using a simple type of learning rule first derived by Girolami (1997) by choosing negentropy as a projection pursuit index. Parameterized probability distributions that have suband super-Gaussian regimes were used to derive a general learning rule that preserves the simple architecture proposed by Bell and Sejnowski (1995), is optimized using the natural gradient by Amari (1998), and uses the stability analysis of Cardoso and Laheld (1996) to switch between sub- and super-Gaussian regimes. We demonstrate that the extended infomax algorithm is able to easily separate 20 sources with a variety of source distributions. Applied to high-dimensional data from electroencephalographic (EEG) recordings, it is effective at separating artifacts such as eye blinks and line noise from weaker electrical signals that arise from sources in the brain. 1
|
808
|
Independent component analysis, a new concept
– Comon
- 1994
|
|
688
|
An information-maximization approach to blind separation and blind deconvolution
– Bell, Sejnowski
- 1995
|
|
372
|
A new learning algorithm for blind signal separation
– Amari, Cichocki, et al.
- 1996
|
|
345
|
Blind separation of sources, Part I: an adaptive algorithm based on a neuromimetic architecture
– Jutten, Hérault
- 1991
|
|
281
|
Equivariant adaptive source separation
– Cardoso, Laheld
- 1996
|
|
264
|
Blind beamforming for non gaussian signals
– Cardoso, Souloumiac
- 1993
|
|
218
|
A fast fixed-point algorithm for independent component analysis
– Hyvärinen, Oja
- 1492
|
|
140
|
Natural gradient works efficiently in learning
– Amari
- 1998
|
|
122
|
Nonlinear neuron in the low noise limit: A factorial code maximizes information transfer
– Nadal, Parga
- 1994
|
|
105
|
A class of neural networks for independent component analysis
– Karhunen, Oja, et al.
- 1997
|
|
96
|
Independent component analysis of electroencephalographic data
– Makeig, Bell, et al.
- 1996
|
|
92
|
Infomax and maximum likelihood for blind source separation
– Cardoso
- 1997
|
|
88
|
A contextsensitive generalization of ICA
– Pearlmutter, Parra
- 1996
|
|
82
|
Source separation in post-nonlinear mixtures
– Taleb, Jutten
- 1999
|
|
72
|
Blind separation of a mixture of independent sources through a quasi-maximum likelihood approach
– Pham, Garat
- 1997
|
|
70
|
An Information-Theoretic Approach to Neural Computing
– Deco, Obradovic
- 1996
|
|
69
|
Robust learning algorithm for blind separation of signals
– Cichocki, Unbehauen, et al.
- 1994
|
|
60
|
Blind source separation of more sources than mixtures using overcomplete representations
– Lee, Lewicki, et al.
- 1999
|
|
56
|
The nonlinear PCA learning rule in independent component analysis
– Oja
- 1997
|
|
48
|
Neural approaches to independent component analysis and source separation
– Karhunen
- 1996
|
|
46
|
Maximum likelihood for blind separation and deconvolution of noisy signals using mixture models
– Moulines, Cardoso, et al.
- 1997
|
|
43
|
source separation – semiparametric statistical approach
– Amari, Cardoso, et al.
- 1997
|
|
38
|
Stability analysis of adaptive blind source separation
– Amari, Chen, et al.
- 1997
|
|
36
|
Spatially independent activity patterns in functional magnetic resonance imaging data during the Stroop color-naming task
– Mckeown, Jung, et al.
- 1998
|
|
36
|
Redundancy reduction and independent component analysis: Conditions on cumulants and adaptive approaches
– Nadal, Parga
- 1997
|
|
32
|
A maximum likelihood approach to nonlinear blind source separation
– Pajunen, Karhunen
- 1997
|
|
31
|
Learning non-linear overcomplete representations for efficient coding
– LEWICKI, SEJNOWSKI
- 1998
|
|
29
|
An alternative perspective on adaptive independent component analysis algorithms
– Girolami
- 1998
|
|
29
|
Multidimensional Density Shaping by Sigmoids
– Roth, Baram
- 1996
|
|
23
|
Feature extraction through LOCOCODE
– Hochreiter, Schmidhuber
- 1999
|
|
21
|
Extended ica removes artifacts from electroencephalographic recordings
– Jung, Humphries, et al.
- 1997
|
|
20
|
Extraction of Independent Signal Sources Using a Deflationary Exploratory Projection Pursuit Network with Lateral Inhibition
– Girolami, Fyfe
- 1997
|
|
18
|
Blind separation of event-related brain responses into independent components
– Makeig, Jung, et al.
- 1997
|
|
16
|
Generalised independent component analysis through unsupervised learning with emergent bussgang properties
– Girolami, Fyfe
- 1997
|
|
14
|
A unifying framework for independent component analysis
– Lee, Girolami, et al.
- 2000
|
|
10
|
Source separation without prior knowledge: the maximum likelihood solution
– Gaeta, Lacoume
- 1990
|
|
9
|
Blind separation of noisy mixtures: An EM algorithm for independent factor analysis. Neural Computation
– Attias
- 1999
|
|
9
|
Self-organizing artificial neural networks for signal separation
– Girolami
- 1997
|
|
9
|
Blind separation of nonlinear mixing models
– Lee, Koehler, et al.
- 1997
|
|
9
|
Faithful representation of separable input distributions
– Lin, Cowan
- 1997
|
|
7
|
Kendall's Advanced Theory of Statistic, 1, Distribution Theory
– Stuart, Ord
- 1987
|
|
7
|
Ica fixed-point algorithm in extraction of artifacts from eeg
– Vigario, Hyvarinen, et al.
- 1996
|
|
7
|
Maximum equalization by entropy maximization and mixture of cumulative distribution functions
– Xu, Cheung, et al.
- 1997
|
|
4
|
Dipole models of eye movements and
– Berg, Scherg
- 1991
|
|
4
|
Blind signal processing: a review
– Cardoso
- 1998
|
|
4
|
Unsupervised adaptive filtering, chapter Entropic contrasts for source separation. Simon Haykin editor
– Cardoso
- 1998
|
|
4
|
Contributions to the mathematical study of evolution
– Pearson
- 1894
|
|
2
|
Changes in the eeg spectrum predict fluctuations in error rate in an auditory vigilance task
– Makeig, Inlow
- 1993
|
|
1
|
submitted 1998c). Unsupervised classification with non-gaussian mixture models using ica
– Lee, Lewicki, et al.
|
|
