Dissimilarity Clustering by Hierarchical Multi-Level Refinement

Brieuc Conan-Guez (1) , Fabrice Rossi (2)

1 LITA - Laboratoire d'Informatique Théorique et Appliquée
2 SAMM - Statistique, Analyse et Modélisation Multidisciplinaire (SAmos-Marin Mersenne)

We introduce in this paper a new way of optimizing the natural extension of the quantization error using in k-means clustering to dissimilarity data. The proposed method is based on hierarchical clustering analysis combined with multi-level heuristic refinement. The method is computationally efficient and achieves better quantization errors than the relational k-means

Domaines: Machine Learning [stat.ML]; Apprentissage [cs.LG]

Brieuc Conan-Guez, Fabrice Rossi. Dissimilarity Clustering by Hierarchical Multi-Level Refinement. 20-th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2012), Apr 2012, Bruges, Belgium. pp.483-488. ⟨hal-00692282⟩