ICA and ISA Using Schweizer-Wolff Measure of Dependence

Abstract

We propose a new algorithm for independent component and independent subspace analysis problems. This algorithm uses a contrast based on the Schweizer-Wolff measure of pairwise dependence, a non-parametric measure based on pairwise ranks of the variables. Our algorithm frequently outperforms state of the art ICA methods in the normal setting, is significantly more robust to outliers in the mixed signals, and performs well even in the presence of noise. Since pairwise dependence is evaluated explicitly, using Cardoso's conjecture, our method can be applied to solve independence subspace analysis (ISA) problems by grouping signals recovered by ICA methods. We provide an extensive empirical evaluation using simulated, sound, and image data.

Cite

Text

Kirshner and Póczos. "ICA and ISA Using Schweizer-Wolff Measure of Dependence." International Conference on Machine Learning, 2008. doi:10.1145/1390156.1390215

Markdown

[Kirshner and Póczos. "ICA and ISA Using Schweizer-Wolff Measure of Dependence." International Conference on Machine Learning, 2008.](https://mlanthology.org/icml/2008/kirshner2008icml-ica/) doi:10.1145/1390156.1390215

BibTeX

@inproceedings{kirshner2008icml-ica,
  title     = {{ICA and ISA Using Schweizer-Wolff Measure of Dependence}},
  author    = {Kirshner, Sergey and Póczos, Barnabás},
  booktitle = {International Conference on Machine Learning},
  year      = {2008},
  pages     = {464-471},
  doi       = {10.1145/1390156.1390215},
  url       = {https://mlanthology.org/icml/2008/kirshner2008icml-ica/}
}