Benchmarking cost-assignment schemes for multi-objective evolutionary algorithms

Koukoulakis, K. and Li, Y. (2000) Benchmarking cost-assignment schemes for multi-objective evolutionary algorithms. In: Cagnoni, S., Poli, R. and Li, Y. (eds.) Real-World Applications of Evolutionary Computing. Series: Lecture notes in computer science, 1803. Springer Berlin Heidelberg: Berlin, pp. 158-167. ISBN 9783540673538 (doi:10.1007/3-540-45561-2_16)

Full text not currently available from Enlighten.

Publisher's URL:


Currently there exist various cost-assignment schemes that perform the necessary scalarization of the objective values when applied to a multi-objective optimization problem. Of course, the final decision depends highly on the nature of the problem but given the multiplicity of the schemes combined with the fact that what the user ultimately needs is a single compromise solution it is evident that elaborating the selection of the method is not a trivial task. This paper intends to address this problem by extending the benchmarks of optimality and reach time given in [1] to mutliobjective optimization problems. A number of existing cost-assignment schemes are evaluated using such benchmarks.

Item Type:Book Sections
Glasgow Author(s) Enlighten ID:Li, Professor Yun
Authors: Koukoulakis, K., and Li, Y.
Subjects:Q Science > Q Science (General)
College/School:College of Science and Engineering > School of Engineering > Systems Power and Energy
Publisher:Springer Berlin Heidelberg

University Staff: Request a correction | Enlighten Editors: Update this record