Title :

Properties of variational estimates of a mixture model for random graphs

Author(s) :

Daudin, Jean-Jacques [Auteur]
Mathématiques et Informatique Appliquées [MIA-Paris]
Célisse, Alain [Auteur]
Mathématiques et Informatique Appliquées [MIA-Paris]
Laboratoire Paul Painlevé - UMR 8524 [LPP]
Gazal, Steven [Auteur]
Mathématiques et Informatique Appliquées [MIA-Paris]
Robin, Stephane [Auteur]
Mathématiques et Informatique Appliquées [MIA-Paris]

Conference title :

ECCS10 European Conference on Complex Systems

City :

Lisbonne

Country :

France

Start date of the conference :

2010-09-13

Book title :

European Conference on Complex Systems 10: programme and Abstracts

Journal title :

European Conference on Complex Systems 10: programme and Abstracts

Publication date :

2010

English keyword(s) :

biological network
mixture model
random graph model
variational inference variational Bayes inference

HAL domain(s) :

Sciences du Vivant [q-bio]

English abstract : [en]

Mixture models for random graphs have a complex depen- dency structure and a likelihood which is not computable even for mod- erate size networks. Variational and variational Bayes techniques are useful approaches for ...
Show more >Mixture models for random graphs have a complex depen- dency structure and a likelihood which is not computable even for mod- erate size networks. Variational and variational Bayes techniques are useful approaches for statistical inference of such complex models but their theoretical properties are not well known. We give a result about the consistency of variational estimates of the parameters of the model and we propose variational Bayes estimates. We compare the accuracy of the two variational methods through simulation studies and show an application to a large Protein-Protein interaction network.Show less >

Language :

Anglais

Peer reviewed article :

Non

Audience :

Non spécifiée

Popular science :

Non

Comment :

Satellite meeting<br/>Satellite meeting

Collections :

• Laboratoire Paul Painlevé - UMR 8524

Source :

Harvested from HAL