Injecting noise for analysing the stability of ICA components

Authors:
Stefan Harmeling;Frank Meinecke;Klaus-Robert Müller
Affiliations:
Fraunhofer First IDA, Kekuléstrasse 7, 12489 Berlin, Germany;Fraunhofer First IDA, Kekuléstrasse 7, 12489 Berlin, Germany;Fraunhofer First IDA, Kekuléstrasse 7, 12489 Berlin, Germany and Department of Computer Science, University of Potsdam, August-Bebel-Strasse 89, 14482 Potsdam, Germany
Venue:
Signal Processing - Special issue on independent components analysis and beyond
Year:
2004

Citing 7
Cited 4

Blind separation of sources, Part 1: an adaptive algorithm based on neuromimetic architecture

Signal Processing
Independent component analysis, a new concept?

Signal Processing - Special issue on higher order statistics
Training with noise is equivalent to Tikhonov regularization

Neural Computation
An information-maximization approach to blind separation and blind deconvolution

Neural Computation
Matrix computations (3rd ed.)

Matrix computations (3rd ed.)
Blind separation of instantaneous mixture of sources via the Gaussian mutual information criterion

Signal Processing
A blind source separation technique using second-order statistics

IEEE Transactions on Signal Processing

A toolbox for model-free analysis of fMRI data

ICA'07 Proceedings of the 7th international conference on Independent component analysis and signal separation
Consistency and asymptotic normality of FastICA and bootstrap FastICA

Signal Processing
Distributional convergence of subspace estimates in FastICA: a bootstrap study

LVA/ICA'12 Proceedings of the 10th international conference on Latent Variable Analysis and Signal Separation
Validation of Nonlinear PCA

Neural Processing Letters

Quantified Score

Hi-index	0.00

Visualization

Abstract

Usually, noise is considered to be destructive. We present a new method that constructively injects noise to assess the reliability and the grouping structure of empirical ICA component estimates. Our method can be viewed as a Monte-Carlo-style approximation of the curvature of some performance measure at the solution. Simulations show that the true root-mean-squared angle distances between the real sources and the source estimates can be approximated well by our method. In a toy experiment, we see that we are also able to reveal the underlying grouping structure of the extracted ICA components. Furthermore, an experiment with fetal ECG data demonstrates that our approach is useful for exploratory data analysis of real-world data.