Abstract
It was recently proven that sets of points maximizing the
hypervolume indicator do not give a good multiplicative approximation
of the Pareto front. We introduce a new “logarithmic
hypervolume indicator” and prove that it achieves a
close-to-optimal multiplicative approximation ratio. This is
experimentally verified on several benchmark functions by
comparing the approximation quality of the multi-objective
covariance matrix evolution strategy (MO-CMA-ES) with
the classic hypervolume indicator and the MO-CMA-ES
with the logarithmic hypervolume indicator.
hypervolume indicator do not give a good multiplicative approximation
of the Pareto front. We introduce a new “logarithmic
hypervolume indicator” and prove that it achieves a
close-to-optimal multiplicative approximation ratio. This is
experimentally verified on several benchmark functions by
comparing the approximation quality of the multi-objective
covariance matrix evolution strategy (MO-CMA-ES) with
the classic hypervolume indicator and the MO-CMA-ES
with the logarithmic hypervolume indicator.
| Originalsprog | Engelsk |
|---|---|
| Titel | Proceedings of the 11th Workshop on Foundations of genetic algorithms : FOGA '11 |
| Redaktører | Hans-Georg Beyer, W. B. Langdon |
| Antal sider | 11 |
| Forlag | Association for Computing Machinery |
| Publikationsdato | 2011 |
| Sider | 81-91 |
| ISBN (Trykt) | 978-1-4503-0633-1 |
| DOI | |
| Status | Udgivet - 2011 |
| Begivenhed | Foundations of genetic algorithms - Schwarzenberg, Østrig Varighed: 5 jan. 2011 → 9 jan. 2011 Konferencens nummer: XI |
Konference
| Konference | Foundations of genetic algorithms |
|---|---|
| Nummer | XI |
| Land/Område | Østrig |
| By | Schwarzenberg |
| Periode | 05/01/2011 → 09/01/2011 |