An EM Stopping Rule for Avoiding Degeneracy in Gaussian-based Clustering with Missing Data - Inria - Institut national de recherche en sciences et technologies du numérique
Communication Dans Un Congrès Année : 2024

An EM Stopping Rule for Avoiding Degeneracy in Gaussian-based Clustering with Missing Data

Résumé

Missing data frequency increases with the growing size of multivariate modern datasets. In Gaussian model-based clustering, the EM algorithm easily takes into account such data but the degeneracy problem is dramatically aggravated during the EM runs: parameter de- generacy is quite slow and also more frequent than with complete data. Consequently, parameter degenerated solutions may be confused with valuable parameter solutions and, in addition, computing time may be wasted through wrong runs. In this work, a simple and low informa- tional condition on the latent partition allows to propose a very simple partition-based stopping rule of EM which shows good behavior on nu- merical experiments.
Fichier principal
Vignette du fichier
SMPS2024.pdf (299.55 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04867801 , version 1 (06-01-2025)

Licence

Identifiants

  • HAL Id : hal-04867801 , version 1

Citer

Christophe Biernacki, Vandewalle Vincent. An EM Stopping Rule for Avoiding Degeneracy in Gaussian-based Clustering with Missing Data. International Conference on Soft Methods in Probability and Statistics, Sep 2024, Salsburg, Austria. ⟨hal-04867801⟩
0 Consultations
0 Téléchargements

Partager

More