Self-organization of predictive representations

J. M. Herrmann, K. Pawelzik, T. Geisel

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract / Description of output

We propose an approach for the development of dynamic representations which are predictive for future sensory inputs. The prediction error allows one to restructure both internal and input connectivity such that, from the initially unstable dynamics of a random network, a reliable behavior is obtained after learning. In particular, we consider the self-organization of connectivities similar to synfire chains (for linear sequences of inputs) or effectively two-dimensional neural layers (for data from an autonomous robot in a maze)
Original languageEnglish
Title of host publicationArtificial Neural Networks, 1999. ICANN 99. Ninth International Conference on (Conf. Publ. No. 470)
Number of pages6
ISBN (Print)0-85296-721-7
Publication statusPublished - 1999

Keywords / Materials (for Non-textual outputs)

  • self-organising feature maps
  • Bayesian networks
  • generalisation
  • input connectivity
  • internal connectivity
  • learning
  • prediction error
  • self-organization


Dive into the research topics of 'Self-organization of predictive representations'. Together they form a unique fingerprint.

Cite this