Consensus Driven Self-Organization: Towards Non Hierarchical Multi-Map Architectures

Noémie Gonnier (1, 2) , Yann Boniface (2) , Hervé Frezza-Buet (1, 2)

1 CentraleSupélec
2 BISCUIT - Bio-Inspired, Situated and Cellular Unconventional Information Technologies

This paper introduces CxSOM, a model to build modular architectures based on self-organizing maps (SOM). An original consensus driven approach enables to adress non-hierarchical architectures where SOMs get organized jointly. The paper aims at showing how the modules are able to store the association between data, and evaluating, by a mutual information criterion, the resulting organization. These results stand as preliminary work to study bigger architectures.

Mots clés: Neural Network Models; Self-Organizing Maps; Multi-Modal Architecture

Domaines: Informatique [cs]

Noémie Gonnier, Yann Boniface, Hervé Frezza-Buet. Consensus Driven Self-Organization: Towards Non Hierarchical Multi-Map Architectures. Communications in Computer and Information Science, Neural Information Processing, ICONIP 2020, pp.526-534, 2020, ⟨10.1007/978-3-030-63823-8_60⟩. ⟨hal-03030518⟩