A New Genetic Algorithm for Set Covering Problems

Aickelin, Uwe (2000) A New Genetic Algorithm for Set Covering Problems. In: Annual Operational Research Conference 42 (OR 42), Swansea, UK.

[img]
Preview
PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
109Kb

Abstract

An indirect genetic algorithm for the non-unicost set covering problem is presented. The algorithm is a two-stage meta-heuristic, which in the past was successfully applied to similar multiple-choice optimisation problems. The two stages of the algorithm are an ‘indirect’ genetic algorithm and a decoder routine. First, the solutions to the problem are encoded as permutations of the rows to be covered, which are subsequently ordered by the genetic algorithm. Fitness assignment is handled by the decoder, which transforms the permutations into actual solutions to the set covering problem. This is done by exploiting both problem structure and problem specific information. However, flexibility is retained by a self-adjusting element within the decoder, which allows adjustments to both the data and to stages within the search process. Computational results are presented.

Item Type:Conference or Workshop Item (Paper)
Schools/Departments:Faculty of Science > School of Computer Science and Information Technology
ID Code:606
Deposited By:Aickelin, Professor Uwe
Deposited On:22 Oct 2007 14:42
Last Modified:22 Oct 2007 14:42

Repository Staff Only: item control page