Probabilistic Topological Map and Binary data

Mustapha Lebbah (1) , C. Chabanon , Sylvie Thiria (2) , Fouad Badran (1)

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

The Self Organizing Map (SOM) proposed by Kohonen is a well known neural model which provides both quantization and clustering of the observation space. In this paper, we adapt the Bernoulli mixture approach to the model of binary topological map and show that using a probabilistic formalism gives rise to better quantization process and classification performances.

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

Mustapha Lebbah, C. Chabanon, Sylvie Thiria, Fouad Badran. Probabilistic Topological Map and Binary data. ICANN2001, Jan 2001, X, France. ⟨hal-01124637⟩