Improved step size adaptation for the MO-CMA-ES

Publikation: Bidrag til bog/antologi/rapportKonferencebidrag i proceedingsForskningfagfællebedømt

Standard

Improved step size adaptation for the MO-CMA-ES. / Voß, T.; Hansen, N.; Igel, Christian.

Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2010). Association for Computing Machinery, 2010. s. 487-494.

Publikation: Bidrag til bog/antologi/rapportKonferencebidrag i proceedingsForskningfagfællebedømt

Harvard

Voß, T, Hansen, N & Igel, C 2010, Improved step size adaptation for the MO-CMA-ES. i Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2010). Association for Computing Machinery, s. 487-494, Gecco 10 Genetic and evolutionary computation conference, Portland, USA, 07/07/2010. https://doi.org/10.1145/1830483.1830573

APA

Voß, T., Hansen, N., & Igel, C. (2010). Improved step size adaptation for the MO-CMA-ES. I Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2010) (s. 487-494). Association for Computing Machinery. https://doi.org/10.1145/1830483.1830573

Vancouver

Voß T, Hansen N, Igel C. Improved step size adaptation for the MO-CMA-ES. I Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2010). Association for Computing Machinery. 2010. s. 487-494 https://doi.org/10.1145/1830483.1830573

Author

Voß, T. ; Hansen, N. ; Igel, Christian. / Improved step size adaptation for the MO-CMA-ES. Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2010). Association for Computing Machinery, 2010. s. 487-494

Bibtex

@inproceedings{8b200c7d25fa49ca9e2261dc4c071ffe,
title = "Improved step size adaptation for the MO-CMA-ES",
abstract = "The multi-objective covariance matrix adaptation evolution strategy (MO-CMA-ES) is an evolutionary algorithm for continuous vector-valued optimization. It combines indicator-based selection based on the contributing hypervolume with the efficient strategy parameter adaptation of the elitist covariance matrix adaptation evolution strategy (CMA-ES). Step sizes (i.e., mutation strengths) are adapted on individual-level using an improved implementation of the 1/5-th success rule. In the original MO-CMA-ES, a mutation is regarded as successful if the offspring ranks better than its parent in the elitist, rank-based selection procedure. In contrast, we propose to regard a mutation as successful if the offspring is selected into the next parental population. This criterion is easier to implement and reduces the computational complexity of the MO-CMA-ES, in particular of its steady-state variant. The new step size adaptation improves the performance of the MO-CMA-ES as shown empirically using a large set of benchmark functions. The new update scheme in general leads to larger step sizes and thereby counteracts premature convergence. The experiments comprise the first evaluation of the MO-CMA-ES for problems with more than two objectives.",
author = "T. Vo{\ss} and N. Hansen and Christian Igel",
year = "2010",
doi = "10.1145/1830483.1830573",
language = "English",
pages = "487--494",
booktitle = "Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2010)",
publisher = "Association for Computing Machinery",
note = "null ; Conference date: 07-07-2010 Through 11-07-2010",

}

RIS

TY - GEN

T1 - Improved step size adaptation for the MO-CMA-ES

AU - Voß, T.

AU - Hansen, N.

AU - Igel, Christian

PY - 2010

Y1 - 2010

N2 - The multi-objective covariance matrix adaptation evolution strategy (MO-CMA-ES) is an evolutionary algorithm for continuous vector-valued optimization. It combines indicator-based selection based on the contributing hypervolume with the efficient strategy parameter adaptation of the elitist covariance matrix adaptation evolution strategy (CMA-ES). Step sizes (i.e., mutation strengths) are adapted on individual-level using an improved implementation of the 1/5-th success rule. In the original MO-CMA-ES, a mutation is regarded as successful if the offspring ranks better than its parent in the elitist, rank-based selection procedure. In contrast, we propose to regard a mutation as successful if the offspring is selected into the next parental population. This criterion is easier to implement and reduces the computational complexity of the MO-CMA-ES, in particular of its steady-state variant. The new step size adaptation improves the performance of the MO-CMA-ES as shown empirically using a large set of benchmark functions. The new update scheme in general leads to larger step sizes and thereby counteracts premature convergence. The experiments comprise the first evaluation of the MO-CMA-ES for problems with more than two objectives.

AB - The multi-objective covariance matrix adaptation evolution strategy (MO-CMA-ES) is an evolutionary algorithm for continuous vector-valued optimization. It combines indicator-based selection based on the contributing hypervolume with the efficient strategy parameter adaptation of the elitist covariance matrix adaptation evolution strategy (CMA-ES). Step sizes (i.e., mutation strengths) are adapted on individual-level using an improved implementation of the 1/5-th success rule. In the original MO-CMA-ES, a mutation is regarded as successful if the offspring ranks better than its parent in the elitist, rank-based selection procedure. In contrast, we propose to regard a mutation as successful if the offspring is selected into the next parental population. This criterion is easier to implement and reduces the computational complexity of the MO-CMA-ES, in particular of its steady-state variant. The new step size adaptation improves the performance of the MO-CMA-ES as shown empirically using a large set of benchmark functions. The new update scheme in general leads to larger step sizes and thereby counteracts premature convergence. The experiments comprise the first evaluation of the MO-CMA-ES for problems with more than two objectives.

U2 - 10.1145/1830483.1830573

DO - 10.1145/1830483.1830573

M3 - Article in proceedings

SP - 487

EP - 494

BT - Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2010)

PB - Association for Computing Machinery

Y2 - 7 July 2010 through 11 July 2010

ER -

ID: 33863100