Automated Design of Collaboration-Based Hybrid Metaheuristics

Hybridization plays a prominent role in bolstering the performance of optimization algorithms (OAs), yet designing efficient hybrid OAs tailored to intricate optimization problems persists as a formidable task. This article introduces a novel top-down methodology for the automated design of hybrid O...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:IEEE transactions on cybernetics Ročník 54; číslo 12; s. 7877 - 7890
Hlavní autori: Wang, Yipeng, Xin, Bin, Liu, Bo, Wang, Qing
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: United States IEEE 01.12.2024
Predmet:
Automated algorithm design (AAD)
Automation
Benchmark testing
Design methodology
hybrid metaheuristics (MHs)
Mathematical models
meta-optimization
Metaheuristics
multifactorial optimization
Multitasking
Taxonomy
ISSN:2168-2267, 2168-2275, 2168-2275
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Hybridization plays a prominent role in bolstering the performance of optimization algorithms (OAs), yet designing efficient hybrid OAs tailored to intricate optimization problems persists as a formidable task. This article introduces a novel top-down methodology for the automated design of hybrid OAs, treating algorithm design as a meta-optimization problem. A general design template for collaboration-based hybrid OAs is developed, integrating a multitude of hybridization strategies for the first time. Besides, a mathematical model is built to formulate the meta-optimization problem of algorithm design. To address the meta-optimization challenge, an improved multifactorial evolutionary algorithm is proposed to automatically design efficient hybrid metaheuristics in a multitasking environment for the given instances with diverse features. To verify the effectiveness of the proposed design methodology, it is applied to the CEC2017 benchmark functions and the binary knapsack problem. Numerical results have demonstrated the feasibility and effectiveness of the proposed methodology for both continuous and combinatorial optimization benchmarks.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:2168-2267
2168-2275
2168-2275
DOI:10.1109/TCYB.2024.3412997