“…Besides non-automatic, but visual, clustering rules , or probabilistic approaches (De Soete & Winsberg, 1993;Séménou, Courcoux, Cardinal, Nicod, & Ouisse, 2007), k-means clustering or hierarchical clustering methods are the most common approaches. The k-means clustering, as the fuzzy clustering (c-means), are partitioning methods that may be considered as non-probabilistic versions of the model-based methods.…”