On the contribution of gene libraries to artificial immune systems

Authors:
Peter Spellward;Tim Kovacs
Affiliations:
University of Bristol, Bristol, UK;University of Bristol, Bristol, UK
Venue:
GECCO '05 Proceedings of the 7th annual conference on Genetic and evolutionary computation
Year:
2005

Citing 5
Cited 1

An algorithm for solving the job-shop problem

Management Science
Adaptation in natural and artificial systems

Adaptation in natural and artificial systems
A Promising Genetic Algorithm Approach to Job-Shop SchedulingRe-Schedulingand Open-Shop Scheduling Problems

Proceedings of the 5th International Conference on Genetic Algorithms
The Evolution of Emergent Organization in Immune System Gene Libraries

Proceedings of the 6th International Conference on Genetic Algorithms
Architecture for an Artificial Immune System

Evolutionary Computation

Investigating a hybrid metaheuristic for job shop rescheduling

ACAL'07 Proceedings of the 3rd Australian conference on Progress in artificial life

Quantified Score

Hi-index	0.00

Visualization

Abstract

Gene libraries have been added to Artificial Immune Systems in analogy to biological immune systems, but to date no careful study of their effect has been made. This work investigates the contribution of gene libraries to Artificial Immune Systems by reproducing and extending an earlier system that used gene libraries. Performance on a job-shop scheduling problem is evaluated empirically with and without gene libraries, and with many different library configurations. We propose that gene libraries encourage diversity in a population of solutions and that the number of components in the gene library parameterises this effect. The number of gene libraries used is found to affect solution fitness and indeed using larger numbers of libraries (and therefore libraries of smaller components) enables higher fitness to be attained. We conclude that gene libraries are likely to be of use in applications where there is a need to maintain the diversity of solutions.