λ-opt neural networks for quadratic assignment problem

Shin Ishii, Hirotaka Niitsuma

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We propose new analog neural approaches to quadratic assignment problems. Our methods are based on an analog version of the λ-opt heuristics, which simultaneously changes assignments for λ elements in a permutation. Since we can take a relatively large λ value, our methods can achieve a middle-range search over the possible solutions, and this helps the system neglect shallow local minima and escape from local minima. Results have shown that our methods are comparable to the present champion algorithms, and for two benchmark problems, they are able to obtain better solutions than the previous champion algorithms.

Original languageEnglish
Title of host publicationIEE Conference Publication
PublisherIEE
Pages115-120
Number of pages6
Volume1
Edition470
ISBN (Print)0852967217
Publication statusPublished - 1999
Externally publishedYes
EventProceedings of the 1999 the 9th International Conference on 'Artificial Neural Networks (ICANN99)' - Edinburgh, UK
Duration: Sep 7 1999Sep 10 1999

Other

OtherProceedings of the 1999 the 9th International Conference on 'Artificial Neural Networks (ICANN99)'
CityEdinburgh, UK
Period9/7/999/10/99

Fingerprint

Neural networks

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Ishii, S., & Niitsuma, H. (1999). λ-opt neural networks for quadratic assignment problem. In IEE Conference Publication (470 ed., Vol. 1, pp. 115-120). IEE.

λ-opt neural networks for quadratic assignment problem. / Ishii, Shin; Niitsuma, Hirotaka.

IEE Conference Publication. Vol. 1 470. ed. IEE, 1999. p. 115-120.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ishii, S & Niitsuma, H 1999, λ-opt neural networks for quadratic assignment problem. in IEE Conference Publication. 470 edn, vol. 1, IEE, pp. 115-120, Proceedings of the 1999 the 9th International Conference on 'Artificial Neural Networks (ICANN99)', Edinburgh, UK, 9/7/99.
Ishii S, Niitsuma H. λ-opt neural networks for quadratic assignment problem. In IEE Conference Publication. 470 ed. Vol. 1. IEE. 1999. p. 115-120
Ishii, Shin ; Niitsuma, Hirotaka. / λ-opt neural networks for quadratic assignment problem. IEE Conference Publication. Vol. 1 470. ed. IEE, 1999. pp. 115-120
@inproceedings{5cc071ea4ce94bb697496287684bc852,
title = "λ-opt neural networks for quadratic assignment problem",
abstract = "We propose new analog neural approaches to quadratic assignment problems. Our methods are based on an analog version of the λ-opt heuristics, which simultaneously changes assignments for λ elements in a permutation. Since we can take a relatively large λ value, our methods can achieve a middle-range search over the possible solutions, and this helps the system neglect shallow local minima and escape from local minima. Results have shown that our methods are comparable to the present champion algorithms, and for two benchmark problems, they are able to obtain better solutions than the previous champion algorithms.",
author = "Shin Ishii and Hirotaka Niitsuma",
year = "1999",
language = "English",
isbn = "0852967217",
volume = "1",
pages = "115--120",
booktitle = "IEE Conference Publication",
publisher = "IEE",
edition = "470",

}

TY - GEN

T1 - λ-opt neural networks for quadratic assignment problem

AU - Ishii, Shin

AU - Niitsuma, Hirotaka

PY - 1999

Y1 - 1999

N2 - We propose new analog neural approaches to quadratic assignment problems. Our methods are based on an analog version of the λ-opt heuristics, which simultaneously changes assignments for λ elements in a permutation. Since we can take a relatively large λ value, our methods can achieve a middle-range search over the possible solutions, and this helps the system neglect shallow local minima and escape from local minima. Results have shown that our methods are comparable to the present champion algorithms, and for two benchmark problems, they are able to obtain better solutions than the previous champion algorithms.

AB - We propose new analog neural approaches to quadratic assignment problems. Our methods are based on an analog version of the λ-opt heuristics, which simultaneously changes assignments for λ elements in a permutation. Since we can take a relatively large λ value, our methods can achieve a middle-range search over the possible solutions, and this helps the system neglect shallow local minima and escape from local minima. Results have shown that our methods are comparable to the present champion algorithms, and for two benchmark problems, they are able to obtain better solutions than the previous champion algorithms.

UR - http://www.scopus.com/inward/record.url?scp=0033341907&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033341907&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0033341907

SN - 0852967217

VL - 1

SP - 115

EP - 120

BT - IEE Conference Publication

PB - IEE

ER -