Component-level aggregation of probabilistic PCA mixtures using variational-Bayes

Pierrick Bruneau; Marc Gelgon; Fabien Picarougne

hal-00476076, version 2

Component-level aggregation of probabilistic PCA mixtures using variational-Bayes

Pierrick Bruneau () ^1, 2, Marc Gelgon () ^1, 2, Fabien Picarougne () ²

Technical Report. This report of an extended version of our ICPR'2010 paper. (2011)

Abstract: This paper proposes a technique for aggregating mixtures of probabilistic principal component analyzers, which are a powerful probabilistic generative model for coping with a high-dimensional, non linear, data set. Aggregation is carried out through Bayesian estimation with a specific prior and an original variational scheme. We demonstrate how such models may be aggregated by accessing model parameters only, rather than original data, which can be advantageous for learning from distributed data sets. Experimental results illustrate the effectiveness of the proposal.

1: ATLAS (INRIA)
INRIA – Université de Nantes
2: Laboratoire d'Informatique de Nantes Atlantique (LINA)
CNRS : UMR6241 – Université de Nantes – École Nationale Supérieure des Mines - Nantes

hal-00476076, version 2
http://hal.archives-ouvertes.fr/hal-00476076
oai:hal.archives-ouvertes.fr:hal-00476076
From: Marc Gelgon <>
Submitted on: Sunday, 20 February 2011 12:23:56
Updated on: Monday, 21 February 2011 13:39:11

See detailed view

Associated documents

PDF :

Export

Bibtex EndNote TEI RefWorks

<?xml version="1.0" encoding="UTF-8"?>
<listBibl xmlns="http://www.tei-c.org/ns/1.0" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML">
<listBibl type="article"><head>Other publications</head>
<biblStruct type="article" xml:id="hal-00476076"><analytic>
<author><persName><forename>Pierrick</forename><surname>Bruneau</surname></persName>
<affiliation>
<orgName type="laboratory">INRIA</orgName>
<orgName type="team">ATLAS</orgName>
<country>FR</country>
<orgName type="institution">INRIA</orgName>
<orgName type="institution">Université de Nantes</orgName>
</affiliation>
<affiliation>
<orgName type="laboratory">LINA</orgName>
<orgName type="team">Laboratoire d'Informatique de Nantes Atlantique</orgName>
<country>FR</country>
<orgName type="institution">CNRS : UMR6241</orgName>
<orgName type="institution">Université de Nantes</orgName>
<orgName type="institution">École Nationale Supérieure des Mines - Nantes</orgName>
</affiliation>
</author>
<author><persName><forename>Marc</forename><surname>Gelgon</surname></persName>
<affiliation>
<orgName type="laboratory">INRIA</orgName>
<orgName type="team">ATLAS</orgName>
<country>FR</country>
<orgName type="institution">INRIA</orgName>
<orgName type="institution">Université de Nantes</orgName>
</affiliation>
<affiliation>
<orgName type="laboratory">LINA</orgName>
<orgName type="team">Laboratoire d'Informatique de Nantes Atlantique</orgName>
<country>FR</country>
<orgName type="institution">CNRS : UMR6241</orgName>
<orgName type="institution">Université de Nantes</orgName>
<orgName type="institution">École Nationale Supérieure des Mines - Nantes</orgName>
</affiliation>
</author>
<author><persName><forename>Fabien</forename><surname>Picarougne</surname></persName>
<affiliation>
<orgName type="laboratory">LINA</orgName>
<orgName type="team">Laboratoire d'Informatique de Nantes Atlantique</orgName>
<country>FR</country>
<orgName type="institution">CNRS : UMR6241</orgName>
<orgName type="institution">Université de Nantes</orgName>
<orgName type="institution">École Nationale Supérieure des Mines - Nantes</orgName>
</affiliation>
</author>
<title type="main" level="a">Component-level aggregation of probabilistic PCA mixtures using variational-Bayes</title></analytic><monogr><imprint><biblScope type="publicationDate">2011-02-00</biblScope></imprint></monogr><idno type="HALid">http://hal.archives-ouvertes.fr/hal-00476076</idno><idno type="HALFile">http://hal.archives-ouvertes.fr/hal-00476076/PDF/techrep.pdf</idno></biblStruct>
</listBibl>
</listBibl>