G
Gilles Celeux
Researcher at French Institute for Research in Computer Science and Automation
Publications - 153
Citations - 12867
Gilles Celeux is an academic researcher from French Institute for Research in Computer Science and Automation. The author has contributed to research in topics: Cluster analysis & Expectation–maximization algorithm. The author has an hindex of 42, co-authored 151 publications receiving 11591 citations. Previous affiliations of Gilles Celeux include Institut de Mathématiques de Toulouse & Agro ParisTech.
Papers
More filters
Journal ArticleDOI
An entropy criterion for assessing the number of clusters in a mixture model
Gilles Celeux,Gilda Soromenho +1 more
TL;DR: In this article, an entropy criterion is proposed to estimate the number of clusters arising from a mixture model, which is derived from a relation linking the likelihood and the classification likelihood of a mixture.
Journal ArticleDOI
Assessing a mixture model for clustering with the integrated completed likelihood
TL;DR: An assessing method of mixture model in a cluster analysis setting with integrated completed likelihood appears to be more robust to violation of some of the mixture model assumptions and it can select a number of dusters leading to a sensible partitioning of the data.
Journal ArticleDOI
Deviance information criteria for missing data models
TL;DR: The deviance information criterion is reassessed for missing data models, testing the behaviour of variousextensions in the cases of mixture and random models.
Journal ArticleDOI
Gaussian parsimonious clustering models
Gilles Celeux,Gérard Govaert +1 more
TL;DR: Methods of optimization to derive the maximum likelihood estimates as well as the practical usefulness of these models are discussed and an application on stellar data which dramatically illustrated the relevance of allowing clusters to have different volumes is illustrated.
Journal ArticleDOI
A Classification EM algorithm for clustering and two stochastic versions
Gilles Celeux,Gérard Govaert +1 more
TL;DR: Two stochastic algorithms are derived from this general Classification EM algorithm, incorporating random perturbations, to reduce the initial-position dependence of the classical optimization clustering algorithms.