Post-nonlinear Independent Component Analysis by Variational Bayesian Learning
Identifieur interne : 000190 ( Main/Exploration ); précédent : 000189; suivant : 000191Post-nonlinear Independent Component Analysis by Variational Bayesian Learning
Auteurs : Alexander Ilin [Finlande] ; Antti Honkela [Finlande]Source :
- Lecture Notes in Computer Science [ 0302-9743 ] ; 2004.
Abstract
Abstract: Post-nonlinear (PNL) independent component analysis(ICA) is a generalisation of ICA where the observations are assumed to have been generated from independent sources by linear mixing followed by component-wise scalar nonlinearities. Most previous PNL ICA algorithms require the post-nonlinearities to be invertible functions. In this paper, we present a variational Bayesian approach to PNL ICA that also works for non-invertible post-nonlinearities. The method is based on a generative model with multi-layer perceptron (MLP) networks to model the post-nonlinearities. Preliminary results with a difficult artificial example are encouraging.
Url:
DOI: 10.1007/978-3-540-30110-3_97
Affiliations:
Links toward previous steps (curation, corpus...)
Le document en format XML
<record><TEI wicri:istexFullTextTei="biblStruct"><teiHeader><fileDesc><titleStmt><title xml:lang="en">Post-nonlinear Independent Component Analysis by Variational Bayesian Learning</title>
<author><name sortKey="Ilin, Alexander" sort="Ilin, Alexander" uniqKey="Ilin A" first="Alexander" last="Ilin">Alexander Ilin</name>
</author>
<author><name sortKey="Honkela, Antti" sort="Honkela, Antti" uniqKey="Honkela A" first="Antti" last="Honkela">Antti Honkela</name>
</author>
</titleStmt>
<publicationStmt><idno type="wicri:source">ISTEX</idno>
<idno type="RBID">ISTEX:B53E814A676C649F357474AD36E84BEAD6C7C628</idno>
<date when="2004" year="2004">2004</date>
<idno type="doi">10.1007/978-3-540-30110-3_97</idno>
<idno type="url">https://api.istex.fr/document/B53E814A676C649F357474AD36E84BEAD6C7C628/fulltext/pdf</idno>
<idno type="wicri:Area/Main/Corpus">000221</idno>
<idno type="wicri:explorRef" wicri:stream="Main" wicri:step="Corpus" wicri:corpus="ISTEX">000221</idno>
<idno type="wicri:Area/Main/Curation">000221</idno>
<idno type="wicri:Area/Main/Exploration">000190</idno>
<idno type="wicri:explorRef" wicri:stream="Main" wicri:step="Exploration">000190</idno>
</publicationStmt>
<sourceDesc><biblStruct><analytic><title level="a" type="main" xml:lang="en">Post-nonlinear Independent Component Analysis by Variational Bayesian Learning</title>
<author><name sortKey="Ilin, Alexander" sort="Ilin, Alexander" uniqKey="Ilin A" first="Alexander" last="Ilin">Alexander Ilin</name>
<affiliation wicri:level="1"><country xml:lang="fr">Finlande</country>
<wicri:regionArea>Neural Networks Research Centre, Helsinki University of Technology, P.O. Box 5400, FI-02015 HUT, Espoo</wicri:regionArea>
<wicri:noRegion>Espoo</wicri:noRegion>
</affiliation>
<affiliation wicri:level="1"><country wicri:rule="url">Finlande</country>
</affiliation>
</author>
<author><name sortKey="Honkela, Antti" sort="Honkela, Antti" uniqKey="Honkela A" first="Antti" last="Honkela">Antti Honkela</name>
<affiliation wicri:level="1"><country xml:lang="fr">Finlande</country>
<wicri:regionArea>Neural Networks Research Centre, Helsinki University of Technology, P.O. Box 5400, FI-02015 HUT, Espoo</wicri:regionArea>
<wicri:noRegion>Espoo</wicri:noRegion>
</affiliation>
<affiliation wicri:level="1"><country wicri:rule="url">Finlande</country>
</affiliation>
</author>
</analytic>
<monogr></monogr>
<series><title level="s">Lecture Notes in Computer Science</title>
<imprint><date>2004</date>
</imprint>
<idno type="ISSN">0302-9743</idno>
<idno type="eISSN">1611-3349</idno>
<idno type="ISSN">0302-9743</idno>
</series>
<idno type="istex">B53E814A676C649F357474AD36E84BEAD6C7C628</idno>
<idno type="DOI">10.1007/978-3-540-30110-3_97</idno>
<idno type="ChapterID">97</idno>
<idno type="ChapterID">Chap97</idno>
</biblStruct>
</sourceDesc>
<seriesStmt><idno type="ISSN">0302-9743</idno>
</seriesStmt>
</fileDesc>
<profileDesc><textClass></textClass>
<langUsage><language ident="en">en</language>
</langUsage>
</profileDesc>
</teiHeader>
<front><div type="abstract" xml:lang="en">Abstract: Post-nonlinear (PNL) independent component analysis(ICA) is a generalisation of ICA where the observations are assumed to have been generated from independent sources by linear mixing followed by component-wise scalar nonlinearities. Most previous PNL ICA algorithms require the post-nonlinearities to be invertible functions. In this paper, we present a variational Bayesian approach to PNL ICA that also works for non-invertible post-nonlinearities. The method is based on a generative model with multi-layer perceptron (MLP) networks to model the post-nonlinearities. Preliminary results with a difficult artificial example are encouraging.</div>
</front>
</TEI>
<affiliations><list><country><li>Finlande</li>
</country>
</list>
<tree><country name="Finlande"><noRegion><name sortKey="Ilin, Alexander" sort="Ilin, Alexander" uniqKey="Ilin A" first="Alexander" last="Ilin">Alexander Ilin</name>
</noRegion>
<name sortKey="Honkela, Antti" sort="Honkela, Antti" uniqKey="Honkela A" first="Antti" last="Honkela">Antti Honkela</name>
<name sortKey="Honkela, Antti" sort="Honkela, Antti" uniqKey="Honkela A" first="Antti" last="Honkela">Antti Honkela</name>
<name sortKey="Ilin, Alexander" sort="Ilin, Alexander" uniqKey="Ilin A" first="Alexander" last="Ilin">Alexander Ilin</name>
</country>
</tree>
</affiliations>
</record>
Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Wicri/France/explor/AussoisV1/Data/Main/Exploration
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000190 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd -nk 000190 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien |wiki= Wicri/France |area= AussoisV1 |flux= Main |étape= Exploration |type= RBID |clé= ISTEX:B53E814A676C649F357474AD36E84BEAD6C7C628 |texte= Post-nonlinear Independent Component Analysis by Variational Bayesian Learning }}
This area was generated with Dilib version V0.6.29. |