Classification croisée et modèles

Issue		RAIRO Oper. Res. Volume 33, Number 4, October-December 1999


Page(s)		525 - 541
DOI		10.1051/ro:1999122

DOI: 10.1051/ro:1999122

RAIRO Rech. Opér. (vol. 33, n $^\circ$ 4, 1999, pp. 525-541)

Classification croisée et modèles

Y. Bencheikh

Institut de Mathematiques, Université Ferhat Abbas de Setif, Setif 19000, Algérie.

Abstract:

The relations between automatic clustering methods and inferentiel statistical models have mostely been studied when the data involves only one set. We propose to study these relations in the case of data involving two sets. We shall look at cross clustering methods as suggested by Govaert [6]; we show that these methods, like the simple clustering methods, can be considered as a clustering approach of a mixture model. We introduce the notion of crossed mixture from a concret example and define the notions of likelihood and associated clustered likelihood. Then, we study the relations which exist between the crossed mixture models and simple models and we show that these relations are completely similar to those which exist between the crossed clustering methods and simple clustering methods.

Résumé:

Les liens existant entre les methodes de classification automatique et les modeles de statistiques inferentielles ont surtout ete etudies lorsque les donnees mettent en jeu un seul ensemble. Nous nous proposons ici de le faire lorsque les donnees mettent en jeu deux ensembles. Nous nous sommes interesses aux methodes de classification croisee proposees par Govaert [6]; nous montrons que ces methodes, comme les methodes de classification simple, peuvent etre considerees, comme une approche classification d'un modele de melange. Nous introduisons la notion de melange croise a partir d'un exemple concret et nous definissons les notions de vraisemblance et de vraisemblance classifiante associees, nous etudions ensuite les liens qui existent entre les modeles de melange croise et les modeles de melange simple et nous montrons que ces liens sont tout a fait analogues a ceux qui existent entre les methodes de classification croisee et les methodes de classification simple.

Keywords: L1 distance, automatic clustering, mixture, cross mixture.

What is OpenURL?

The OpenURL standard is a protocol for transmission of metadata describing the resource that you wish to access. An OpenURL link contains article metadata and directs it to the OpenURL server of your choice. The OpenURL server can provide access to the resource and also offer complementary services (specific search engine, export of references...). The OpenURL link can be generated by different means.

If your librarian has set up your subscription with an OpenURL resolver, OpenURL links appear automatically on the abstract pages.
You can define your own OpenURL resolver with your EDPS Account. In this case your choice will be given priority over that of your library.
You can use an add-on for your browser (Firefox or I.E.) to display OpenURL links on a page (see http://www.openly.com/openurlref/). You should disable this module if you wish to use the OpenURL server that you or your library have defined.