Using a randomised iterative improvement algorithm with composite neighbourhood structures for the university course timetabling problem

Salwani Abdullah, Edmund K. Burke, Barry McCollum

Research output: Contribution to journalArticle

56 Citations (Scopus)

Abstract

The course timetabling problem deals with the assignment of a set of courses to specific timeslots and rooms within a working week subject to a variety of hard and soft constraints. Solutions which satisfy the hard constraints are called feasible. The goal is to satisfy as many of the soft constraints as possible whilst constructing a feasible schedule. In this paper, we present a composite neighbourhood structure with a randomised iterative improvement algorithm. This algorithm always accepts an improved solution and a worse solution is accepted with a certain probability. The algorithm is tested over eleven benchmark datasets (representing one large, five medium and five small problems). The results demonstrate that our approach is able to produce solutions that have lower penalty on all the small problems and two of the medium problems when compared against other techniques from the literature. However, in the case of the medium problems, this is at the expense of significantly increased computational time.

Original languageEnglish
Pages (from-to)153-169
Number of pages17
JournalOperations Research/ Computer Science Interfaces Series
Volume39
Publication statusPublished - 2007
Externally publishedYes

Fingerprint

Composite structures
Timetabling

ASJC Scopus subject areas

  • Management Science and Operations Research
  • Computer Science(all)

Cite this

@article{5c3d85b061d04a6bbdee088e266f3634,
title = "Using a randomised iterative improvement algorithm with composite neighbourhood structures for the university course timetabling problem",
abstract = "The course timetabling problem deals with the assignment of a set of courses to specific timeslots and rooms within a working week subject to a variety of hard and soft constraints. Solutions which satisfy the hard constraints are called feasible. The goal is to satisfy as many of the soft constraints as possible whilst constructing a feasible schedule. In this paper, we present a composite neighbourhood structure with a randomised iterative improvement algorithm. This algorithm always accepts an improved solution and a worse solution is accepted with a certain probability. The algorithm is tested over eleven benchmark datasets (representing one large, five medium and five small problems). The results demonstrate that our approach is able to produce solutions that have lower penalty on all the small problems and two of the medium problems when compared against other techniques from the literature. However, in the case of the medium problems, this is at the expense of significantly increased computational time.",
author = "Salwani Abdullah and Burke, {Edmund K.} and Barry McCollum",
year = "2007",
language = "English",
volume = "39",
pages = "153--169",
journal = "Operations Research/ Computer Science Interfaces Series",
issn = "1387-666X",
publisher = "Springer Science + Business Media",

}

TY - JOUR

T1 - Using a randomised iterative improvement algorithm with composite neighbourhood structures for the university course timetabling problem

AU - Abdullah, Salwani

AU - Burke, Edmund K.

AU - McCollum, Barry

PY - 2007

Y1 - 2007

N2 - The course timetabling problem deals with the assignment of a set of courses to specific timeslots and rooms within a working week subject to a variety of hard and soft constraints. Solutions which satisfy the hard constraints are called feasible. The goal is to satisfy as many of the soft constraints as possible whilst constructing a feasible schedule. In this paper, we present a composite neighbourhood structure with a randomised iterative improvement algorithm. This algorithm always accepts an improved solution and a worse solution is accepted with a certain probability. The algorithm is tested over eleven benchmark datasets (representing one large, five medium and five small problems). The results demonstrate that our approach is able to produce solutions that have lower penalty on all the small problems and two of the medium problems when compared against other techniques from the literature. However, in the case of the medium problems, this is at the expense of significantly increased computational time.

AB - The course timetabling problem deals with the assignment of a set of courses to specific timeslots and rooms within a working week subject to a variety of hard and soft constraints. Solutions which satisfy the hard constraints are called feasible. The goal is to satisfy as many of the soft constraints as possible whilst constructing a feasible schedule. In this paper, we present a composite neighbourhood structure with a randomised iterative improvement algorithm. This algorithm always accepts an improved solution and a worse solution is accepted with a certain probability. The algorithm is tested over eleven benchmark datasets (representing one large, five medium and five small problems). The results demonstrate that our approach is able to produce solutions that have lower penalty on all the small problems and two of the medium problems when compared against other techniques from the literature. However, in the case of the medium problems, this is at the expense of significantly increased computational time.

UR - http://www.scopus.com/inward/record.url?scp=57849127576&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=57849127576&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:57849127576

VL - 39

SP - 153

EP - 169

JO - Operations Research/ Computer Science Interfaces Series

JF - Operations Research/ Computer Science Interfaces Series

SN - 1387-666X

ER -