Comparing Generic Parameter Controllers for EAs

G. Karafotias, M. Hoogendoorn

Research output: Chapter in Book / Report / Conference proceedingConference contributionAcademicpeer-review

Abstract

Parameter controllers for Evolutionary Algorithms (EAs) deal with adjusting parameter values during an evolutionary run. Many ad hoc approaches have been presented for parameter control, but few generic parameter controllers exist and, additionally, no comparisons or in depth analyses of these generic controllers are available in literature. This paper presents an extensive comparison of such generic parameter control methods, including a number of novel controllers based on reinforcement learning which are introduced here. We conducted experiments with different EAs and test problems in an one-off setting, i.e. relatively long runs with controllers used out-of-the-box with no tailoring to the problem at hand. Results reveal several interesting insights regarding the effectiveness of parameter control, the niche applications/EAs, the effect of continuous treatment of parameters and the influence of noise and randomness on control.
LanguageEnglish
Title of host publicationIEEE Symposium Series on Computational Intelligence (SSCI '14)
PublisherIEEE
Pages46-53
DOIs
Publication statusPublished - 2014

Fingerprint

Evolutionary algorithms
Controllers
Reinforcement learning
Experiments

Cite this

Karafotias, G., & Hoogendoorn, M. (2014). Comparing Generic Parameter Controllers for EAs. In IEEE Symposium Series on Computational Intelligence (SSCI '14) (pp. 46-53). IEEE. https://doi.org/10.1109/FOCI.2014.7007806
Karafotias, G. ; Hoogendoorn, M. / Comparing Generic Parameter Controllers for EAs. IEEE Symposium Series on Computational Intelligence (SSCI '14). IEEE, 2014. pp. 46-53
@inproceedings{5a5ca3d570204a3d908a3b382309b6da,
title = "Comparing Generic Parameter Controllers for EAs",
abstract = "Parameter controllers for Evolutionary Algorithms (EAs) deal with adjusting parameter values during an evolutionary run. Many ad hoc approaches have been presented for parameter control, but few generic parameter controllers exist and, additionally, no comparisons or in depth analyses of these generic controllers are available in literature. This paper presents an extensive comparison of such generic parameter control methods, including a number of novel controllers based on reinforcement learning which are introduced here. We conducted experiments with different EAs and test problems in an one-off setting, i.e. relatively long runs with controllers used out-of-the-box with no tailoring to the problem at hand. Results reveal several interesting insights regarding the effectiveness of parameter control, the niche applications/EAs, the effect of continuous treatment of parameters and the influence of noise and randomness on control.",
author = "G. Karafotias and M. Hoogendoorn",
year = "2014",
doi = "10.1109/FOCI.2014.7007806",
language = "English",
pages = "46--53",
booktitle = "IEEE Symposium Series on Computational Intelligence (SSCI '14)",
publisher = "IEEE",

}

Karafotias, G & Hoogendoorn, M 2014, Comparing Generic Parameter Controllers for EAs. in IEEE Symposium Series on Computational Intelligence (SSCI '14). IEEE, pp. 46-53. https://doi.org/10.1109/FOCI.2014.7007806

Comparing Generic Parameter Controllers for EAs. / Karafotias, G.; Hoogendoorn, M.

IEEE Symposium Series on Computational Intelligence (SSCI '14). IEEE, 2014. p. 46-53.

Research output: Chapter in Book / Report / Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Comparing Generic Parameter Controllers for EAs

AU - Karafotias, G.

AU - Hoogendoorn, M.

PY - 2014

Y1 - 2014

N2 - Parameter controllers for Evolutionary Algorithms (EAs) deal with adjusting parameter values during an evolutionary run. Many ad hoc approaches have been presented for parameter control, but few generic parameter controllers exist and, additionally, no comparisons or in depth analyses of these generic controllers are available in literature. This paper presents an extensive comparison of such generic parameter control methods, including a number of novel controllers based on reinforcement learning which are introduced here. We conducted experiments with different EAs and test problems in an one-off setting, i.e. relatively long runs with controllers used out-of-the-box with no tailoring to the problem at hand. Results reveal several interesting insights regarding the effectiveness of parameter control, the niche applications/EAs, the effect of continuous treatment of parameters and the influence of noise and randomness on control.

AB - Parameter controllers for Evolutionary Algorithms (EAs) deal with adjusting parameter values during an evolutionary run. Many ad hoc approaches have been presented for parameter control, but few generic parameter controllers exist and, additionally, no comparisons or in depth analyses of these generic controllers are available in literature. This paper presents an extensive comparison of such generic parameter control methods, including a number of novel controllers based on reinforcement learning which are introduced here. We conducted experiments with different EAs and test problems in an one-off setting, i.e. relatively long runs with controllers used out-of-the-box with no tailoring to the problem at hand. Results reveal several interesting insights regarding the effectiveness of parameter control, the niche applications/EAs, the effect of continuous treatment of parameters and the influence of noise and randomness on control.

U2 - 10.1109/FOCI.2014.7007806

DO - 10.1109/FOCI.2014.7007806

M3 - Conference contribution

SP - 46

EP - 53

BT - IEEE Symposium Series on Computational Intelligence (SSCI '14)

PB - IEEE

ER -

Karafotias G, Hoogendoorn M. Comparing Generic Parameter Controllers for EAs. In IEEE Symposium Series on Computational Intelligence (SSCI '14). IEEE. 2014. p. 46-53 https://doi.org/10.1109/FOCI.2014.7007806