Do Hypervolume Regressions Hinder EMOA Performance? Surprise and Relief
Decreases in dominated hypervolume w.r.t a fixed reference point for the (μ + 1)-SMS-EMOA are able to appear. We examine the impact of these decreases and different reference point handling techniques by providing four different algorithmic variants for selection. In addition, we show that yet further decreases can occur due to numerical instabilities that were previously not being expected. Fortunately, our findings do indicate that all detected decreases do not have a negative effect on the overall performance.
KeywordsEMO hypervolume decreases reference point handling numerical instabilities
Unable to display preview. Download preview PDF.
- 5.Deb, K., Thiele, L., Laumanns, M., Zitzler, E.: Scalable Multi-Objective Optimization Test Problems. In: Congress on Evolutionary Computation (CEC 2002), pp. 825–830. IEEE Press (2002)Google Scholar
- 7.Judt, L., Mersmann, O., Naujoks, B.: Non-monotonicity of Obtained Hypervolume in 1-greedy S-Metric Selection. Journal of Multi-Criteria Decision Analysis (2011), maanvs03.gm.fh-koeln.de/webpub/CIOPReports.d/Judt11a.d/ (accepted for publication, preprint)
- 9.Mersmann, O.: emoa: Evolutionary Multiobjective Optimization Algorithms (2011), http://CRAN.R-project.org/package=emoa (R package version 0.4-8)