Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric tests

<p>Non-parametric tests can determine the better of two stochastic optimization algorithms when benchmarking results are ordinal—like the final fitness values of multiple trials—but for many benchmarks, a trial can also terminate once it reaches a prespecified target value. In such cases, both...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Kenneth V. Price (17877002) (author)
مؤلفون آخرون: Abhishek Kumar (156479) (author), P.N. Suganthan (16518528) (author)
منشور في: 2023
الموضوعات:
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1864513528421941248
author Kenneth V. Price (17877002)
author2 Abhishek Kumar (156479)
P.N. Suganthan (16518528)
author2_role author
author
author_facet Kenneth V. Price (17877002)
Abhishek Kumar (156479)
P.N. Suganthan (16518528)
author_role author
dc.creator.none.fl_str_mv Kenneth V. Price (17877002)
Abhishek Kumar (156479)
P.N. Suganthan (16518528)
dc.date.none.fl_str_mv 2023-04-01T00:00:00Z
dc.identifier.none.fl_str_mv 10.1016/j.swevo.2023.101287
dc.relation.none.fl_str_mv https://figshare.com/articles/journal_contribution/Trial-based_dominance_for_comparing_both_the_speed_and_accuracy_of_stochastic_optimizers_with_standard_non-parametric_tests/25117088
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Mathematical sciences
Applied mathematics
Benchmarking
Two-variable non-parametric tests
Evolutionary algorithms
Dominance
Stochastic optimization
Numerical optimization
Mann-Whitney test
dc.title.none.fl_str_mv Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric tests
dc.type.none.fl_str_mv Text
Journal contribution
info:eu-repo/semantics/publishedVersion
text
contribution to journal
description <p>Non-parametric tests can determine the better of two stochastic optimization algorithms when benchmarking results are ordinal—like the final fitness values of multiple trials—but for many benchmarks, a trial can also terminate once it reaches a prespecified target value. In such cases, both the time that a trial takes to reach the target value (or not) and its final fitness value characterize its outcome. This paper describes how trial-based dominance can totally order this two-variable dataset of outcomes so that traditional non-parametric methods can determine the better of two algorithms when one is faster, but less accurate than the other, i.e. when neither algorithm dominates. After describing trial-based dominance, we outline its benefits. We subsequently review other attempts to compare stochastic optimizers, before illustrating our method with the Mann-Whitney U test. Simulations demonstrate that “U-scores” are much more effective than dominance when tasked with identifying the better of two algorithms. We validate U-scores by having them determine the winners of the CEC 2022 competition on single objective, bound-constrained numerical optimization.</p><h2>Other Information</h2> <p> Published in: Swarm and Evolutionary Computation<br> License: <a href="http://creativecommons.org/licenses/by/4.0/" target="_blank">http://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1016/j.swevo.2023.101287" target="_blank">https://dx.doi.org/10.1016/j.swevo.2023.101287</a></p>
eu_rights_str_mv openAccess
id Manara2_5d301428a933a31830cd6f542c71f032
identifier_str_mv 10.1016/j.swevo.2023.101287
network_acronym_str Manara2
network_name_str Manara2
oai_identifier_str oai:figshare.com:article/25117088
publishDate 2023
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric testsKenneth V. Price (17877002)Abhishek Kumar (156479)P.N. Suganthan (16518528)Mathematical sciencesApplied mathematicsBenchmarkingTwo-variable non-parametric testsEvolutionary algorithmsDominanceStochastic optimizationNumerical optimizationMann-Whitney test<p>Non-parametric tests can determine the better of two stochastic optimization algorithms when benchmarking results are ordinal—like the final fitness values of multiple trials—but for many benchmarks, a trial can also terminate once it reaches a prespecified target value. In such cases, both the time that a trial takes to reach the target value (or not) and its final fitness value characterize its outcome. This paper describes how trial-based dominance can totally order this two-variable dataset of outcomes so that traditional non-parametric methods can determine the better of two algorithms when one is faster, but less accurate than the other, i.e. when neither algorithm dominates. After describing trial-based dominance, we outline its benefits. We subsequently review other attempts to compare stochastic optimizers, before illustrating our method with the Mann-Whitney U test. Simulations demonstrate that “U-scores” are much more effective than dominance when tasked with identifying the better of two algorithms. We validate U-scores by having them determine the winners of the CEC 2022 competition on single objective, bound-constrained numerical optimization.</p><h2>Other Information</h2> <p> Published in: Swarm and Evolutionary Computation<br> License: <a href="http://creativecommons.org/licenses/by/4.0/" target="_blank">http://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1016/j.swevo.2023.101287" target="_blank">https://dx.doi.org/10.1016/j.swevo.2023.101287</a></p>2023-04-01T00:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.1016/j.swevo.2023.101287https://figshare.com/articles/journal_contribution/Trial-based_dominance_for_comparing_both_the_speed_and_accuracy_of_stochastic_optimizers_with_standard_non-parametric_tests/25117088CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/251170882023-04-01T00:00:00Z
spellingShingle Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric tests
Kenneth V. Price (17877002)
Mathematical sciences
Applied mathematics
Benchmarking
Two-variable non-parametric tests
Evolutionary algorithms
Dominance
Stochastic optimization
Numerical optimization
Mann-Whitney test
status_str publishedVersion
title Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric tests
title_full Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric tests
title_fullStr Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric tests
title_full_unstemmed Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric tests
title_short Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric tests
title_sort Trial-based dominance for comparing both the speed and accuracy of stochastic optimizers with standard non-parametric tests
topic Mathematical sciences
Applied mathematics
Benchmarking
Two-variable non-parametric tests
Evolutionary algorithms
Dominance
Stochastic optimization
Numerical optimization
Mann-Whitney test