Discovering the phase of a dynamical system from a stream of partial observations with a multi-map self-organizing architecture

Bassem Khouzam (1) , Hervé Frezza-Buet (2, 3)

1 IMS : Information, Multimodalité & Signal
2 SUPELEC-Campus Metz
3 Georgia Tech Lorraine [Metz]

This paper presents a self-organizing architecture made of several maps, implementing a recurrent neural network to cope with partial observations of the phase of some dynamical system. The purpose of self-organization is to set up a distributed representation of the actual phase, although the observations received from the system are ambiguous (i.e. the same observation may correspond to distinct phases). The setting up of such a representation is illustrated by experiments, and then the paper concludes on extensions toward adaptive state representations for partially observable Markovian decision processes.

Domaines: Calcul parallèle, distribué et partagé [cs.DC]

Bassem Khouzam, Hervé Frezza-Buet. Discovering the phase of a dynamical system from a stream of partial observations with a multi-map self-organizing architecture. COGNITIVE 2011, Sep 2011, Rome, Italy. pp.19-24. ⟨hal-00652308⟩