License
when quoting this document, please refer to the following
URN: urn:nbn:de:0030-drops-14213
URL: http://drops.dagstuhl.de/opus/volltexte/2008/1421/

Van der Velde, Frank ; de Kamps, Marc

The role of recurrent networks in neural architectures of grounded cognition: learning of control

pdf-format:
Dokument 1.pdf (200 KB)


Abstract

Recurrent networks have been used as neural models of language processing, with mixed results. Here, we discuss the role of recurrent networks in a neural architecture of grounded cognition. In particular, we discuss how the control of binding in this architecture can be learned. We trained a simple recurrent network (SRN) and a feedforward network (FFN) for this task. The results show that information from the architecture is needed as input for these networks to learn control of binding. Thus, both control systems are recurrent. We found that the recurrent system consisting of the architecture and an SRN or an FFN as a ‘core’ can learn basic (but recursive) sentence structures. Problems with control of binding arise when the system with the SRN is tested on number of new sentence structures. In contrast, control of binding for these structures succeeds with the FFN. Yet, for some structures with (unlimited) embeddings, difficulties arise due to dynamical binding conflicts in the architecture itself. In closing, we discuss potential future developments of the architecture presented here.

BibTeX - Entry

@InProceedings{vandervelde_et_al:DSP:2008:1421,
  author =	{Frank Van der Velde and Marc de Kamps},
  title =	{The role of recurrent networks in neural architectures of grounded cognition: learning of control},
  booktitle =	{Recurrent Neural Networks- Models, Capacities, and Applications},
  year =	{2008},
  editor =	{Luc De Raedt and Barbara Hammer and Pascal Hitzler and Wolfgang Maass},
  number =	{08041},
  series =	{Dagstuhl Seminar Proceedings},
  ISSN =	{1862-4405},
  publisher =	{Internationales Begegnungs- und Forschungszentrum f{\"u}r Informatik (IBFI), Schloss Dagstuhl, Germany},
  address =	{Dagstuhl, Germany},
  URL =		{http://drops.dagstuhl.de/opus/volltexte/2008/1421},
  annote =	{Keywords: Grounded representations, binding control, combinatorial structures, neural architecture, recurrent network, learning}
}

Keywords: Grounded representations, binding control, combinatorial structures, neural architecture, recurrent network, learning
Seminar: 08041 - Recurrent Neural Networks- Models, Capacities, and Applications
Issue date: 2008
Date of publication: 15.04.2008


DROPS-Home | Fulltext Search | Imprint Published by LZI