Detailed Information

Cited 0 time in webofscience Cited 0 time in scopus
Metadata Downloads

Surrogate-Assisted Hybrid-Model Estimation of Distribution Algorithm for Mixed-Variable Hyperparameters Optimization in Convolutional Neural Networks

Full metadata record
DC Field Value Language
dc.contributor.authorLi, Jian-Yu-
dc.contributor.authorZhan, Zhi-Hui-
dc.contributor.authorXu, Jin-
dc.contributor.authorKwong, Sam-
dc.contributor.authorZhang, Jun-
dc.date.accessioned2024-05-02T02:30:25Z-
dc.date.available2024-05-02T02:30:25Z-
dc.date.issued2023-05-
dc.identifier.issn2162-237X-
dc.identifier.issn2162-2388-
dc.identifier.urihttps://scholarworks.bwise.kr/erica/handle/2021.sw.erica/118933-
dc.description.abstractThe performance of a convolutional neural network (CNN) heavily depends on its hyperparameters. However, finding a suitable hyperparameters configuration is difficult, challenging, and computationally expensive due to three issues, which are 1) the mixed-variable problem of different types of hyperparameters; 2) the large-scale search space of finding optimal hyperparameters; and 3) the expensive computational cost for evaluating candidate hyperparameters configuration. Therefore, this article focuses on these three issues and proposes a novel estimation of distribution algorithm (EDA) for efficient hyperparameters optimization, with three major contributions in the algorithm design. First, a hybrid-model EDA is proposed to efficiently deal with the mixed-variable difficulty. The proposed algorithm uses a mixed-variable encoding scheme to encode the mixed-variable hyperparameters and adopts an adaptive hybrid-model learning (AHL) strategy to efficiently optimize the mixed-variables. Second, an orthogonal initialization (OI) strategy is proposed to efficiently deal with the challenge of large-scale search space. Third, a surrogate-assisted multi-level evaluation (SME) method is proposed to reduce the expensive computational cost. Based on the above, the proposed algorithm is named surrogate-assisted hybrid-model EDA (SHEDA). For experimental studies, the proposed SHEDA is verified on widely used classification benchmark problems, and is compared with various state-of-the-art methods. Moreover, a case study on aortic dissection (AD) diagnosis is carried out to evaluate its performance. Experimental results show that the proposed SHEDA is very effective and efficient for hyperparameters optimization, which can find a satisfactory hyperparameters configuration for the CIFAR10, CIFAR100, and AD diagnosis with only 0.58, 0.97, and 1.18 GPU days, respectively.-
dc.format.extent15-
dc.language영어-
dc.language.isoENG-
dc.publisherIEEE Computational Intelligence Society-
dc.titleSurrogate-Assisted Hybrid-Model Estimation of Distribution Algorithm for Mixed-Variable Hyperparameters Optimization in Convolutional Neural Networks-
dc.typeArticle-
dc.publisher.location미국-
dc.identifier.doi10.1109/TNNLS.2021.3106399-
dc.identifier.scopusid2-s2.0-85115675797-
dc.identifier.wosid000732122600001-
dc.identifier.bibliographicCitationIEEE Transactions on Neural Networks and Learning Systems, v.34, no.5, pp 2338 - 2352-
dc.citation.titleIEEE Transactions on Neural Networks and Learning Systems-
dc.citation.volume34-
dc.citation.number5-
dc.citation.startPage2338-
dc.citation.endPage2352-
dc.type.docTypeArticle-
dc.description.isOpenAccessN-
dc.description.journalRegisteredClassscie-
dc.description.journalRegisteredClassscopus-
dc.relation.journalResearchAreaComputer Science-
dc.relation.journalResearchAreaEngineering-
dc.relation.journalWebOfScienceCategoryComputer Science, Artificial Intelligence-
dc.relation.journalWebOfScienceCategoryComputer Science, Hardware & Architecture-
dc.relation.journalWebOfScienceCategoryComputer Science, Theory & Methods-
dc.relation.journalWebOfScienceCategoryEngineering, Electrical & Electronic-
dc.subject.keywordPlusPARTICLE SWARM OPTIMIZATION-
dc.subject.keywordPlusEVOLUTIONARY ALGORITHM-
dc.subject.keywordAuthorOptimization-
dc.subject.keywordAuthorConvolutional neural networks-
dc.subject.keywordAuthorEstimation-
dc.subject.keywordAuthorComputational modeling-
dc.subject.keywordAuthorBrain modeling-
dc.subject.keywordAuthorProbabilistic logic-
dc.subject.keywordAuthorFeature extraction-
dc.subject.keywordAuthorAortic dissection (AD) diagnosis-
dc.subject.keywordAuthorconvolutional neural network (CNN)-
dc.subject.keywordAuthordeep learning-
dc.subject.keywordAuthorestimation of distribution algorithm (EDA)-
dc.subject.keywordAuthorevolutionary computation (EC)-
dc.subject.keywordAuthorhybrid model-
dc.subject.keywordAuthorhyperparameters optimization-
dc.subject.keywordAuthormixed variable-
dc.identifier.urlhttps://ieeexplore.ieee.org/document/9540902-
Files in This Item
Go to Link
Appears in
Collections
COLLEGE OF ENGINEERING SCIENCES > SCHOOL OF ELECTRICAL ENGINEERING > 1. Journal Articles

qrcode

Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.

Related Researcher

Researcher ZHANG, Jun photo

ZHANG, Jun
ERICA 공학대학 (SCHOOL OF ELECTRICAL ENGINEERING)
Read more

Altmetrics

Total Views & Downloads

BROWSE