Skip to main content

On learning and energy-entropy dependence in recurrent and nonrecurrent signed networks

Journal of Statistical Physics volume 1, pages 319–350 (1969)Cite this article

Abstract

Learning of patterns by neural networks obeying general rules of sensory transduction and of converting membrane potentials to spiking frequencies is considered. Any finite number of cellsA can sample a pattern playing on any finite number of cells ∇ without causing irrevocable sampling bias ifA = ℬ orA ∩ ℬ =

. Total energy transfer from inputs ofA to outputs of ℬ depends on the entropy of the input distribution. Pattern completion on recall trials can occur without destroying perfect memory even ifA = ℬ by choosing the signal thresholds sufficiently large. The mathematical results are global limit and oscillation theorems for a class of nonlinear functional-differential systems.

This is a preview of subscription content, access via your institution.

Access options

Buy single article

Instant access to the full article PDF.

USD 39.95

Price includes VAT (USA)
Tax calculation will be finalised during checkout.

References

  1. S. Grossberg, “Some networks that can learn, remember, and reproduce any number of complicated space-time patterns (I),”J. Math. Mechanics (July 1969).

  2. S. Grossberg, “Some networks that can learn, remember, and reproduce any number of complicated space-time patterns (II),”SIAM J. Applied Math., submitted for publication.

  3. S. Grossberg, “How do overlapping nonrecurrent synaptic fields learn without mutual interference?” (in preparation).

  4. R. B. Livingston, “Brain mechanisms in conditioning and learning,” in:Neurosciences Research Symposium Summaries, F. O. Schmittet al., eds. (MIT Press, Cambridge, Massachusetts, 1967), Vol. 2.

    Google Scholar 

  5. S. Grossberg, “On neural pattern discrimination,”J, Theoret. Biol., submitted for publication.

  6. S. Grossberg, “On learning, information, lateral inhibition, and transmitters,”Math. Biosci. 4:255–310 (1969).

    Google Scholar 

  7. S. Grossberg, “A prediction theory for some nonlinear functional-differential equations (II),”J. Math. Anal. Appl. 22:490–522 (1968).

    Google Scholar 

  8. P. Hartman,Ordinary Differential Equations (John Wiley and Sons, New York, 1964).

    Google Scholar 

  9. S. Grossberg, “A prediction theory for some nonlinear functional-differential equations (I),”J. Math. Anal. Appl. 21:643–694 (1968).

    Google Scholar 

  10. S. Grossberg, “Some nonlinear networks capable of learning a spatial pattern of arbitrary complexity,”Proc. Natl. Acad. Sci. (U.S.) 59:368–372 (1968).

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Massachusetts Institute of Technology, Cambridge, Massachusetts

    Stephen Grossberg

Authors
  1. Stephen Grossberg
    View author publications

    You can also search for this author in PubMed Google Scholar

Additional information

The preparation of this work was supported in part by the National Science Foundation (GP 9003), the Office of Naval Research (N00014-67-A-024-OQ16), and the A.P. Sloan Foundation.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Grossberg, S. On learning and energy-entropy dependence in recurrent and nonrecurrent signed networks. J Stat Phys 1, 319–350 (1969). https://doi.org/10.1007/BF01007484

Download citation

  • Received:

  • Issue Date:

  • DOI: https://doi.org/10.1007/BF01007484

Key words

  • learning
  • stimulus sampling
  • nonlinear difference-differential equations
  • global limits and oscillations
  • flows on signed networks
  • functional-differential systems
  • energy-entropy dependence
  • pattern completion
  • recurrent and nonrecurrent anatomy
  • sensory transduction rules
  • ratio limit theorems