Source Separation and Density Estimation by Faithful Equivariant SOM
Abstract
We couple the tasks of source separation and density estimation by extracting the local geometrical structure of distributions ob(cid:173) tained from mixtures of statistically independent sources. Our modifications of the self-organizing map (SOM) algorithm results in purely digital learning rules which perform non-parametric his(cid:173) togram density estimation. The non-parametric nature of the sep(cid:173) aration allows for source separation of non-linear mixtures. An anisotropic coupling is introduced into our SOM with the role of aligning the network locally with the independent component con(cid:173) tours. This approach provides an exact verification condition for source separation with no prior on the source distributions.
Cite
Text
Lin et al. "Source Separation and Density Estimation by Faithful Equivariant SOM." Neural Information Processing Systems, 1996.Markdown
[Lin et al. "Source Separation and Density Estimation by Faithful Equivariant SOM." Neural Information Processing Systems, 1996.](https://mlanthology.org/neurips/1996/lin1996neurips-source/)BibTeX
@inproceedings{lin1996neurips-source,
title = {{Source Separation and Density Estimation by Faithful Equivariant SOM}},
author = {Lin, Juan K. and Cowan, Jack D. and Grier, David G.},
booktitle = {Neural Information Processing Systems},
year = {1996},
pages = {536-542},
url = {https://mlanthology.org/neurips/1996/lin1996neurips-source/}
}