Unbounded population MO-CMA-ES for the bi-objective BBOB test suite

Oswin Krause; Tobias Glasmachers; Nikolaus Hansen; Christian Igel

doi:10.1145/2908961.2931699

Unbounded population MO-CMA-ES for the bi-objective BBOB test suite

Oswin Krause, Tobias Glasmachers, Nikolaus Hansen, Christian Igel

Datalogisk Institut

13 Citationer (Scopus)

Abstract

The unbounded population multi-objective covariance matrix adaptation evolution strategy∼(UP-MO-CMA-ES) aims at maximizing the total hypervolume covered by all evaluated points. It adds all non-dominated solutions found to its population and employs Gaussian mutations with adaptive covariance matrices to also solve ill-conditioned problems. A novel recombination operator adapts the covariance matrices to point along the Pareto front. The UP-MO-CMA-ES is combined with a parallel exploration strategy and empirically evaluated on the bi-objective BBOB-biobj benchmark problems. Results show that the algorithm can reliably solve ill-conditioned problems as well as weakly-structured problems. However, it is less suited for the rugged multi-modal objective functions in the benchmark.

Originalsprog	Engelsk
Titel	Proceedings of the 2016 Genetic and Evolutionary Computation Conference Companion
Antal sider	8
Forlag	Association for Computing Machinery
Publikationsdato	20 jul. 2016
Sider	1177-1184
ISBN (Trykt)	978-1-4503-4323-7
DOI	https://doi.org/10.1145/2908961.2931699
Status	Udgivet - 20 jul. 2016
Begivenhed	Genetic and Evolutionary Computation Conference - Denver, USA Varighed: 20 jul. 2016 → 24 jul. 2016

Konference

Konference	Genetic and Evolutionary Computation Conference
Land/Område	USA
By	Denver
Periode	20/07/2016 → 24/07/2016

Adgang til dokumentet

10.1145/2908961.2931699

Citationsformater

Unbounded population MO-CMA-ES for the bi-objective BBOB test suite. / Krause, Oswin; Glasmachers, Tobias; Hansen, Nikolaus et al.
Proceedings of the 2016 Genetic and Evolutionary Computation Conference Companion. Association for Computing Machinery, 2016. s. 1177-1184.

Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › peer review

Krause, O, Glasmachers, T, Hansen, N & Igel, C 2016, Unbounded population MO-CMA-ES for the bi-objective BBOB test suite. i Proceedings of the 2016 Genetic and Evolutionary Computation Conference Companion. Association for Computing Machinery, s. 1177-1184, Genetic and Evolutionary Computation Conference, Denver, USA, 20/07/2016. https://doi.org/10.1145/2908961.2931699

@inproceedings{093e1d9190b94e35b45ede18516985a6,

title = "Unbounded population MO-CMA-ES for the bi-objective BBOB test suite",

abstract = "The unbounded population multi-objective covariance matrix adaptation evolution strategy∼(UP-MO-CMA-ES) aims at maximizing the total hypervolume covered by all evaluated points. It adds all non-dominated solutions found to its population and employs Gaussian mutations with adaptive covariance matrices to also solve ill-conditioned problems. A novel recombination operator adapts the covariance matrices to point along the Pareto front. The UP-MO-CMA-ES is combined with a parallel exploration strategy and empirically evaluated on the bi-objective BBOB-biobj benchmark problems. Results show that the algorithm can reliably solve ill-conditioned problems as well as weakly-structured problems. However, it is less suited for the rugged multi-modal objective functions in the benchmark.",

author = "Oswin Krause and Tobias Glasmachers and Nikolaus Hansen and Christian Igel",

year = "2016",

month = jul,

day = "20",

doi = "10.1145/2908961.2931699",

language = "English",

isbn = "978-1-4503-4323-7",

pages = "1177--1184",

booktitle = "Proceedings of the 2016 Genetic and Evolutionary Computation Conference Companion",

publisher = "Association for Computing Machinery",

note = "Genetic and Evolutionary Computation Conference ; Conference date: 20-07-2016 Through 24-07-2016",

}

TY - GEN

T1 - Unbounded population MO-CMA-ES for the bi-objective BBOB test suite

AU - Krause, Oswin

AU - Glasmachers, Tobias

AU - Hansen, Nikolaus

AU - Igel, Christian

PY - 2016/7/20

Y1 - 2016/7/20

N2 - The unbounded population multi-objective covariance matrix adaptation evolution strategy∼(UP-MO-CMA-ES) aims at maximizing the total hypervolume covered by all evaluated points. It adds all non-dominated solutions found to its population and employs Gaussian mutations with adaptive covariance matrices to also solve ill-conditioned problems. A novel recombination operator adapts the covariance matrices to point along the Pareto front. The UP-MO-CMA-ES is combined with a parallel exploration strategy and empirically evaluated on the bi-objective BBOB-biobj benchmark problems. Results show that the algorithm can reliably solve ill-conditioned problems as well as weakly-structured problems. However, it is less suited for the rugged multi-modal objective functions in the benchmark.

AB - The unbounded population multi-objective covariance matrix adaptation evolution strategy∼(UP-MO-CMA-ES) aims at maximizing the total hypervolume covered by all evaluated points. It adds all non-dominated solutions found to its population and employs Gaussian mutations with adaptive covariance matrices to also solve ill-conditioned problems. A novel recombination operator adapts the covariance matrices to point along the Pareto front. The UP-MO-CMA-ES is combined with a parallel exploration strategy and empirically evaluated on the bi-objective BBOB-biobj benchmark problems. Results show that the algorithm can reliably solve ill-conditioned problems as well as weakly-structured problems. However, it is less suited for the rugged multi-modal objective functions in the benchmark.

U2 - 10.1145/2908961.2931699

DO - 10.1145/2908961.2931699

M3 - Article in proceedings

SN - 978-1-4503-4323-7

SP - 1177

EP - 1184

BT - Proceedings of the 2016 Genetic and Evolutionary Computation Conference Companion

PB - Association for Computing Machinery

T2 - Genetic and Evolutionary Computation Conference

Y2 - 20 July 2016 through 24 July 2016

ER -

Unbounded population MO-CMA-ES for the bi-objective BBOB test suite

Abstract

Konference

Adgang til dokumentet

Fingeraftryk

Citationsformater