Capacity of the upstart algorithm

Ansgar H L West, David Saad

Research output: Chapter in Book/Report/Conference proceedingChapter


The storage capacity of multilayer networks with overlapping receptive fields is investigated for a constructive algorithm within a one-step replica symmetry breaking (RSB) treatment. We find that the storage capacity increases logarithmically with the number of hidden units <span class='mathrm'>K</span> without saturating the Mitchison-Durbin bound. The slope of the logarithmic increase decays exponentionally with the stability with which the patterns have been stored.
Original languageEnglish
Title of host publicationMathematics of Neural Networks: Models, Algorithms and Applications
EditorsStephen W. Ellacott, John C. Mason, Iain J. Anderson
Place of PublicationOxford
Number of pages7
ISBN (Print)0-7923-9933-1
Publication statusPublished - 1997

Bibliographical note

The original publication is available at


  • algorithms
  • design
  • experimentation
  • human factors
  • measurement
  • performance
  • reliability
  • security
  • theory
  • werification

Fingerprint Dive into the research topics of 'Capacity of the upstart algorithm'. Together they form a unique fingerprint.

  • Cite this

    West, A. H. L., & Saad, D. (1997). Capacity of the upstart algorithm. In S. W. Ellacott, J. C. Mason, & I. J. Anderson (Eds.), Mathematics of Neural Networks: Models, Algorithms and Applications (pp. 372-378). Kluwer.