Sorting with a neural net

P.J. Zwietering, E.H.L. Aarts, J. Wessels

Research output: Book/ReportReportAcademic

36 Downloads (Pure)

Abstract

In this paper a three-layered perceptron is derived that exactly solves the problem of sorting n numbers. The perceptron configuration has O(n^2) nodes and can be viewed as a "neural" implementation of the parallel enumerative sorting algorithm developed by Preparata. Furthermore, it is shown that the problem of sorting n numbers cannot be solved exactly by a two-layered perceptron. Keywords: Sorting, Combinatorial Optimization, Neural Network, Multi-Layered Perceptron, Classification.
Original languageEnglish
Place of PublicationEindhoven
PublisherTechnische Universiteit Eindhoven
Number of pages17
Publication statusPublished - 1991

Publication series

NameMemorandum COSOR
Volume9110
ISSN (Print)0926-4493

Fingerprint Dive into the research topics of 'Sorting with a neural net'. Together they form a unique fingerprint.

  • Cite this

    Zwietering, P. J., Aarts, E. H. L., & Wessels, J. (1991). Sorting with a neural net. (Memorandum COSOR; Vol. 9110). Technische Universiteit Eindhoven.