Advertisement

Autoassociative memory with high storage capacity

  • Paulo J. L. Adeodato
  • John G. Taylor
Oral Presentations: Theory Theory I: Associative Memory
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1112)

Abstract

The general neural unit (GNU) [1] is known for its high storage capacity as an autoassociative memory. The exponential increase in its storage capacity with the number of inputs per neuron is far greater than the linear growth in the famous Hopfield network [2]. This paper shows that the GNU attains an even higher capacity with the use of pyramids of neurons instead of single neurons as its nodes. The paper also shows that the storage capacity/cost ratio increases, giving further support to this node upgrade. This analysis combines the modular approach for storage capacity assessment of pyramids [3] and of GNUs [4].

Keywords

Autoassociative memory storage capacity RAM-based neural networks general neural unit (GNU) pyramidal architecture 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Aleksander I.: IEE Comp. & Cont. Eng. J., 1, 1990, 259–265.Google Scholar
  2. 2.
    Wong K. Y. M., Sherrington D.: Europhysics Letters, 7 (3), 1988, 197–202.Google Scholar
  3. 3.
    Adeodato P. J. L., Taylor J. G.: Neural Network World, 6, (3), 1996, 241–249.Google Scholar
  4. 4.
    Adeodato P. J. L., Taylor J. G.: In: Proc. of Weightless Neural Networks Workshop, D. L. Bisset (ed.), Canterbury, UK, Sep. 1995, 103–110.Google Scholar
  5. 5.
    Aleksander I.: In: Neural Computing Architectures, I. Aleksander (ed.), 1989, North Oxford Academic Publishers Ltd., London, UK.Google Scholar
  6. 6.
    Wong K. Y. M., Sherrington D.: J. Phys., 22, 1989, 2233–2263.Google Scholar
  7. 7.
    Lucy J.: Electronics Letters, 27 (10), 1991, 799–800.Google Scholar
  8. 8.
    Braga A. P.: Electronics Letters, 30 (1), 1994, 55–56.Google Scholar
  9. 9.
    Rower R., Morciniec M.: Neural Computation, 8, (3), 1996, 629–642.Google Scholar
  10. 10.
    Wong K. Y. M., Sherrington D.: Europhysics Letters, 10 (5), 1989, 419–425.Google Scholar
  11. 11.
    Adeodato P. J. L., Taylor J. G.: In Proc. of ICANN'95, F. Fogelman-Soulié and P. Gallinari (eds.), Paris, France, Oct., 1995, 607–612.Google Scholar
  12. 12.
    Guan Y., Clarkson T., Taylor J. G., Gorse D.: Neural Networks, 7, (3), 1994, 523–538.Google Scholar
  13. 13.
    Filho E. C. D. B. C., Bisset D. L., Fairhurst M. C.: Patt. Rec. Lett., 12 (3), 1992, 131–138.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1996

Authors and Affiliations

  • Paulo J. L. Adeodato
    • 1
  • John G. Taylor
    • 1
  1. 1.Mathematics DepartmentKing's College LondonLondonUK

Personalised recommendations