Abstract
We study the storage capacity for temporal sequences of random patterns in networks with arbitrary delays, evolving under parallel dynamics. For sequences with a common periodT, made up of patterns which remain constant for Δ time steps, couplings with delays τ=Δ·k−1, wherek is integer, are particularly important since they aretuned to the “rhythm” of the sequences. For networks with tuned delays only, we calculate the optimal storage capacity along the lines of Gardner [1] and find identical results to corresponding static cases, whereas untuned couplings induce several complications. For Δ=2, we consider networks with finite fractions 1−a of untuned couplings, additionally weighted in strength by a parameter λ with respect to the tuned couplings. For\(\lambda ^2 (1 - a)<< \) we already find a pronounced decrease of the optimal storage capacity compared to the network where the fraction (1−a) of untuned connections was cut. Thus for optimal error-free storage, the untuned couplings should be switched off. On the other hand, if errors are allowed and the couplings are chosen by a Hebbian prescription, the untuned couplings turn out to be useful, if the fractiona of tuned couplings exceeds a certain critical value, and the weight parameter λ can then be optimized with respect to the storage capacity.
Similar content being viewed by others
References
Gardner, E.: J. Phys. A21, 257 (1988)
Domany, E., van Hemmen, J.L., Schulten, K. (eds.): Physics of neural networks. Berlin, Heidelberg, New York, Tokyo: Springer 1990
Amit, D.J.: Modeling brain function, or the world of attractor neural networks. Cambridge: Cambridge University Press 1989 and ref. therein
Amit, D.J., Gutfreund, H., Sompolinsky, H.: Ann. Phys.173, 30 (1987)
Derrida, B., Gardner, E., Zippelius, A.: Europhys. Lett.4, 167 (1987)
Hopfield, J.J.: Proc. Natl. Acad. Sci. USA79, 2554 (1982); Kohonen, T.: Selforganization and associative memory, 2nd eds. Berlin, Heidelberg, New York: Springer 1987; Personnaz, L., Guyon, I., Dreyfus, G.: J. Phys. (Paris)46, 359 (1985); Diederich, S., Opper, M.: Phys. Rev. Lett.58, 949 (1987); Pöppel, G., Krey, U.: Europhys. Lett.4, 979 (1987); Krauth, W., Mezared, M.: J. Phys. A20, L 745 (1987); Anlauf, J.K., Biehl, M.: Europhys. Lett.10, 687 (1989)
Kühn, R., van Hemmen, J.L.: Preprint (to appear in: Physics of Neural Networks, Domany, E., van Hemmen, J.L., Schulten, K. (eds.). Berlin, Heidelberg, New York, Tokyo: Springer 1990
Bauer, K., Krey, U.: Z. Phys. B — Condensed Matter79, 461 (1990)
Dehaene, S., Changeux, J.P., Nadal, J.P.: Proc. Natl. Acad. Sci. USA84, 2727 (1987)
Kerszberg, M., Zippelius, A.: Preprint (1988) Phys. Scr. (to appear)
Herz, A., Sulzer, B., Kühn, R., van Hemmen, J.L.: Europhys. Lett.7, 663 (1988); Biol. Cybern.60, 457 (1989)
Nadal, J.-P.: Proceedings of the Conference on Measures of Complexity, Rome, Oct. 1987. In: Lecture Notes in Physics. Peliti, L., Vulpiani, A. (eds.), Vol. 314, p. 35. Berlin, Heidelberg, New York, Tokyo: Springer 1988
Li, Z., Herz, A.V.M.: In: Neural networks. Proceedings of the XI. Sitges Conf. 1990, Garrido, L. (ed.) Berlin, Heidelberg, New York, Tokyo: Springer 1990
Keppler, T.B., Abott, L.F.: J. Phys. (Paris)49, 1657 (1988)
Pöppel, G.: PhD thesis, Regensburg 1990
Krey, U., Pöppel, G.: Z. Phys. B — Condensed Matter76, 513 (1989)
Fassnacht, C.: Diploma thesis; Fassnacht, C., Zippelius, A.: Preprint Göttingen 1990
Bouten, M., Engel, A., Komoda, A., Serneels, R.: J. Phys. A23, 4643 (1990)
Gardner, E., Derrida, B.: J. Phys. A21, 271 (1988)
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Bauer, K., Krey, U. On the storage capacity for temporal pattern sequences in networks with delays. Z. Physik B - Condensed Matter 84, 131–141 (1991). https://doi.org/10.1007/BF01453766
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF01453766