A New Hierarchical Clustering Method using Topological Map

Meziane Yacoub (1) , Ndeye Niang Keita (1) , Fouad Badran (1) , Sylvie Thiria (2)

1 CEDRIC - Centre d'études et de recherche en informatique et communications
2 LODYC - Laboratoire d'océanographie dynamique et de climatologie

We present a new hierarchical clustering criteria which can be applied to data set. This is done after generating an initial partition byusing a Topological Self Organizing Map. This criteria contains two terms which take into account two different errors simultaneously: the square error of the entire clustering (as the Ward criteria) and the topological structure given by the Self Organizing Map. A parameter T allows to control the corresponding in uence of these two terms. Results on simulated data are presented which show the effect of this criteria for different values of T.

Domaines: Informatique [cs]; Statistiques [math.ST]

Meziane Yacoub, Ndeye Niang Keita, Fouad Badran, Sylvie Thiria. A New Hierarchical Clustering Method using Topological Map. 10th International Symposium on Applied Stochastique Models and Data Analysis (AMSDA2001), 2001, Compiègne, France. ⟨hal-01124631⟩