The capacity and attractor basins of associative memory models

N. Davey; S. Hunt

doi:10.1007/BFb0098189

The capacity and attractor basins of associative memory models

N. Davey, S. Hunt

Research output: Contribution to journal › Article › peer-review

6 Citations (Scopus)

58 Downloads (Pure)

Abstract

The performance characteristics of five variants of the Hopfield network are examined. Two performance metrics are used: memory capacity, and a measure of the size of basins of attraction. We find that the posttraining adjustment of processor thresholds has, at best, little or no effect on performance, and at worst a significant negative effect. The adoption of a local learning rule can, however, give rise to significant performance gains.

Original language	English
Pages (from-to)	330-339
Journal	Lecture Notes in Computer Science (LNCS)
Volume	1606
DOIs	https://doi.org/10.1007/BFb0098189
Publication status	Published - 1999

Access to Document

10.1007/BFb0098189

901699.pdf

Cite this

@article{97d335382e4349b5a1152c2dd068a5aa,

title = "The capacity and attractor basins of associative memory models",

abstract = "The performance characteristics of five variants of the Hopfield network are examined. Two performance metrics are used: memory capacity, and a measure of the size of basins of attraction. We find that the posttraining adjustment of processor thresholds has, at best, little or no effect on performance, and at worst a significant negative effect. The adoption of a local learning rule can, however, give rise to significant performance gains.",

author = "N. Davey and S. Hunt",

note = "The original publication is available at www.springerlink.com . Copyright Springer ",

year = "1999",

doi = "10.1007/BFb0098189",

language = "English",

volume = "1606",

pages = "330--339",

journal = "Lecture Notes in Computer Science (LNCS)",

issn = "0302-9743",

publisher = "Springer Nature ",

}

TY - JOUR

T1 - The capacity and attractor basins of associative memory models

AU - Davey, N.

AU - Hunt, S.

N1 - The original publication is available at www.springerlink.com . Copyright Springer

PY - 1999

Y1 - 1999

N2 - The performance characteristics of five variants of the Hopfield network are examined. Two performance metrics are used: memory capacity, and a measure of the size of basins of attraction. We find that the posttraining adjustment of processor thresholds has, at best, little or no effect on performance, and at worst a significant negative effect. The adoption of a local learning rule can, however, give rise to significant performance gains.

AB - The performance characteristics of five variants of the Hopfield network are examined. Two performance metrics are used: memory capacity, and a measure of the size of basins of attraction. We find that the posttraining adjustment of processor thresholds has, at best, little or no effect on performance, and at worst a significant negative effect. The adoption of a local learning rule can, however, give rise to significant performance gains.

U2 - 10.1007/BFb0098189

DO - 10.1007/BFb0098189

M3 - Article

SN - 0302-9743

VL - 1606

SP - 330

EP - 339

JO - Lecture Notes in Computer Science (LNCS)

JF - Lecture Notes in Computer Science (LNCS)

ER -

The capacity and attractor basins of associative memory models

Abstract

Access to Document

Fingerprint

Cite this