Unified framework for the adaptive operator selection of discrete parameters

Mudita Sharma; Manuel López-Ibáñez; Dimitar Lubomirov Kazakov

Unified framework for the adaptive operator selection of discrete parameters

Mudita Sharma, Manuel López-Ibáñez, Dimitar Lubomirov Kazakov

Computer Science

Research output: Working paper › Preprint

Abstract

We conduct an exhaustive survey of adaptive selection of operators (AOS) in Evolutionary Algorithms (EAs). We simplified the AOS structure by adding more components to the framework to built upon the existing categorisation of AOS methods. In addition to simplifying, we looked at the commonality among AOS methods from literature to generalise them. Each component is presented with a number of alternative choices, each represented with a formula. We make three sets of comparisons. First, the methods from literature are tested on the BBOB test bed with their default hyper parameters. Second, the hyper parameters of these methods are tuned using an offline configurator known as IRACE. Third, for a given set of problems, we use IRACE to select the best combination of components and tune their hyper parameters.

Original language	English
Number of pages	36
Publication status	Published - 12 May 2020

Publication series

Name	arXiv

Keywords

online tuning
IRACE
differential evolution

Access to Document

https://arxiv.org/abs/2005.05613Licence: Other

Cite this

@techreport{426557630c5747ae82274d33f4a577b9,

title = "Unified framework for the adaptive operator selection of discrete parameters",

abstract = "We conduct an exhaustive survey of adaptive selection of operators (AOS) in Evolutionary Algorithms (EAs). We simplified the AOS structure by adding more components to the framework to built upon the existing categorisation of AOS methods. In addition to simplifying, we looked at the commonality among AOS methods from literature to generalise them. Each component is presented with a number of alternative choices, each represented with a formula. We make three sets of comparisons. First, the methods from literature are tested on the BBOB test bed with their default hyper parameters. Second, the hyper parameters of these methods are tuned using an offline configurator known as IRACE. Third, for a given set of problems, we use IRACE to select the best combination of components and tune their hyper parameters.",

keywords = "online tuning, IRACE, differential evolution",

author = "Mudita Sharma and Manuel L{\'o}pez-Ib{\'a}{\~n}ez and Kazakov, {Dimitar Lubomirov}",

year = "2020",

month = may,

day = "12",

language = "English",

series = "arXiv",

type = "WorkingPaper",

}

TY - UNPB

T1 - Unified framework for the adaptive operator selection of discrete parameters

AU - Sharma, Mudita

AU - López-Ibáñez, Manuel

AU - Kazakov, Dimitar Lubomirov

PY - 2020/5/12

Y1 - 2020/5/12

N2 - We conduct an exhaustive survey of adaptive selection of operators (AOS) in Evolutionary Algorithms (EAs). We simplified the AOS structure by adding more components to the framework to built upon the existing categorisation of AOS methods. In addition to simplifying, we looked at the commonality among AOS methods from literature to generalise them. Each component is presented with a number of alternative choices, each represented with a formula. We make three sets of comparisons. First, the methods from literature are tested on the BBOB test bed with their default hyper parameters. Second, the hyper parameters of these methods are tuned using an offline configurator known as IRACE. Third, for a given set of problems, we use IRACE to select the best combination of components and tune their hyper parameters.

AB - We conduct an exhaustive survey of adaptive selection of operators (AOS) in Evolutionary Algorithms (EAs). We simplified the AOS structure by adding more components to the framework to built upon the existing categorisation of AOS methods. In addition to simplifying, we looked at the commonality among AOS methods from literature to generalise them. Each component is presented with a number of alternative choices, each represented with a formula. We make three sets of comparisons. First, the methods from literature are tested on the BBOB test bed with their default hyper parameters. Second, the hyper parameters of these methods are tuned using an offline configurator known as IRACE. Third, for a given set of problems, we use IRACE to select the best combination of components and tune their hyper parameters.

KW - online tuning

KW - IRACE

KW - differential evolution

M3 - Preprint

T3 - arXiv

BT - Unified framework for the adaptive operator selection of discrete parameters

ER -