Scaling properties of neural networks for job-shop scheduling

Simon Y. Foo, Yoshiyasu Takefuji, Harold Szu

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

This paper investigates the scaling properties of neural networks for solving job-shop scheduling problems. Specifically, the Tank-Hopfield linear programming network is modified to solve mixed integer linear programming with the addition of step-function amplifiers. Using a linear energy function, our approach avoids the traditional problems associated with most Hopfield networks using quadratic energy functions. Although our approach requires more hardware (in terms of processing elements and resistive interconnects) than a recent approach by Zhou et al. [2], the neurons in the modified Tank-Hopfield network do not perform extensive calculations unlike those described by Zhou et al.

Original languageEnglish
Pages (from-to)79-91
Number of pages13
JournalNeurocomputing
Volume8
Issue number1
DOIs
Publication statusPublished - 1995
Externally publishedYes

Fingerprint

Linear Programming
Neural networks
Linear programming
Neurons
Hardware
Processing
Job shop scheduling

Keywords

  • Mixed integer linear programing
  • Optimization
  • Scheduling
  • Tank-Hopfield neural networks

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Cognitive Neuroscience
  • Cellular and Molecular Neuroscience

Cite this

Scaling properties of neural networks for job-shop scheduling. / Foo, Simon Y.; Takefuji, Yoshiyasu; Szu, Harold.

In: Neurocomputing, Vol. 8, No. 1, 1995, p. 79-91.

Research output: Contribution to journalArticle

Foo, Simon Y. ; Takefuji, Yoshiyasu ; Szu, Harold. / Scaling properties of neural networks for job-shop scheduling. In: Neurocomputing. 1995 ; Vol. 8, No. 1. pp. 79-91.
@article{a0445df217484c16b46a1ebf940f9be0,
title = "Scaling properties of neural networks for job-shop scheduling",
abstract = "This paper investigates the scaling properties of neural networks for solving job-shop scheduling problems. Specifically, the Tank-Hopfield linear programming network is modified to solve mixed integer linear programming with the addition of step-function amplifiers. Using a linear energy function, our approach avoids the traditional problems associated with most Hopfield networks using quadratic energy functions. Although our approach requires more hardware (in terms of processing elements and resistive interconnects) than a recent approach by Zhou et al. [2], the neurons in the modified Tank-Hopfield network do not perform extensive calculations unlike those described by Zhou et al.",
keywords = "Mixed integer linear programing, Optimization, Scheduling, Tank-Hopfield neural networks",
author = "Foo, {Simon Y.} and Yoshiyasu Takefuji and Harold Szu",
year = "1995",
doi = "10.1016/0925-2312(94)00011-5",
language = "English",
volume = "8",
pages = "79--91",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - Scaling properties of neural networks for job-shop scheduling

AU - Foo, Simon Y.

AU - Takefuji, Yoshiyasu

AU - Szu, Harold

PY - 1995

Y1 - 1995

N2 - This paper investigates the scaling properties of neural networks for solving job-shop scheduling problems. Specifically, the Tank-Hopfield linear programming network is modified to solve mixed integer linear programming with the addition of step-function amplifiers. Using a linear energy function, our approach avoids the traditional problems associated with most Hopfield networks using quadratic energy functions. Although our approach requires more hardware (in terms of processing elements and resistive interconnects) than a recent approach by Zhou et al. [2], the neurons in the modified Tank-Hopfield network do not perform extensive calculations unlike those described by Zhou et al.

AB - This paper investigates the scaling properties of neural networks for solving job-shop scheduling problems. Specifically, the Tank-Hopfield linear programming network is modified to solve mixed integer linear programming with the addition of step-function amplifiers. Using a linear energy function, our approach avoids the traditional problems associated with most Hopfield networks using quadratic energy functions. Although our approach requires more hardware (in terms of processing elements and resistive interconnects) than a recent approach by Zhou et al. [2], the neurons in the modified Tank-Hopfield network do not perform extensive calculations unlike those described by Zhou et al.

KW - Mixed integer linear programing

KW - Optimization

KW - Scheduling

KW - Tank-Hopfield neural networks

UR - http://www.scopus.com/inward/record.url?scp=0029294452&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029294452&partnerID=8YFLogxK

U2 - 10.1016/0925-2312(94)00011-5

DO - 10.1016/0925-2312(94)00011-5

M3 - Article

VL - 8

SP - 79

EP - 91

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

IS - 1

ER -