Combining Mixture Models and Spectral Clustering for Data Partitioning - Université Grenoble Alpes Accéder directement au contenu
Communication Dans Un Congrès Année : 2020

Combining Mixture Models and Spectral Clustering for Data Partitioning

Résumé

Gaussian Mixture Models are widely used nowadays, thanks to the simplicity and efficiency of the Expectation-Maximization algorithm. However, determining the optimal number of components is tricky and, in the context of data partitioning, may differ from the actual number of clusters. We propose to apply a post-processing step by means of Spectral Clustering: it allows a clever merging of similar Gaussians thanks to the Bhattacharyya distance so that clusters of any shape are automatically discovered. The proposed method shows a significant improvement compared to the classical Gaussian Mixture clustering approach and promising results against well-known partitioning algorithms with respect to the number of parameters.
Fichier principal
Vignette du fichier
Muzeau2020 - Combining Mixture Models and Spectral Clustering for Data Partitioning.pdf (1.07 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03106804 , version 1 (12-01-2021)

Identifiants

Citer

Julien Muzeau, Maria Oliver-Parera, Patricia Ladret, Pascal Bertolino. Combining Mixture Models and Spectral Clustering for Data Partitioning. ICIAR 2020 - 17th International Conference on Image Analysis and Recognition, Jun 2020, Póvoa de Varzim, Portugal. pp.63-75, ⟨10.1007/978-3-030-50516-5_6⟩. ⟨hal-03106804⟩
84 Consultations
449 Téléchargements

Altmetric

Partager

Gmail Mastodon Facebook X LinkedIn More