Dynamic Cell Structures (DCS) represent a family of artificial neural architectures suited both for unsupervised and supervised learning. They belong to the recently [Martinetz94] introduced class of Topology Representing Networks (TRN) which build perfectly topology preserving feature maps. DCS employ a modified Kohonen learning rule in conjunction with competitive Hebbian learning. The Kohonen type learning rule serves to adjust the synaptic weight vectors while Hebbian learning establishes a dynamic lateral connection structure between the units reflecting the topology of the feature manifold. In case of supervised learning, i.e. function approximation, each neural unit implements a Radial Basis Function, and an additional layer of linear output units adjusts according to a delta-rule. DCS is the first RBF-based approximation scheme attempting to concurrently learn and utilize a perfectly topology preserving map for improved performance. Simulations on a selection of CMU-Benchmarks indicate that the DCS idea applied to the Growing Cell Structure algorithm [Fritzke93b] leads to an efficient and elegant algorithm that can beat conventional models on similar tasks.


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Dynamic Cell Structures: Radial Basis Function Networks with Perfect Topology Preservation


    Beteiligte:
    Bruske, Jörg (Autor:in) / Sommer, Gerald (Autor:in)

    Erscheinungsdatum :

    1994-01-01


    Medientyp :

    Paper


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Klassifikation :

    DDC:    0 / 004 / 629



    Topology Optimization Using Hyper Radial Basis Function Network

    Apte, Aditya P. / Wang, Bo Ping | AIAA | 2008



    3D Topology Optimization Using Hyper Radial Basis Function Network

    Apte, A. / Wang, B. | British Library Conference Proceedings | 2009



    AIAA-2006-1814 Topology Optimization Using Hyper Radial Basis Function Network

    Wang, B. / Apte, A. / American Institute of Aeronautics and Astronautics | British Library Conference Proceedings | 2006