Modified differential evolution: a greedy random strategy for genetic recombination

Over the past three decades Evolutionary Algorithms have emerged as a powerful mechanism for finding solutions to large and complex problems. A promising new evolutionary algorithm known as Differential Evolution (DE) was recently introduced and has garnered significant attention in the research lit...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Omega (Oxford) Ročník 33; číslo 3; s. 255 - 265
Hlavní autoři: Bergey, Paul K., Ragsdale, Cliff
Médium: Journal Article
Jazyk:angličtina
Vydáno: Exeter Elsevier Ltd 01.06.2005
Elsevier
Elsevier Science Publishers
Pergamon Press Inc
Edice:Omega
Témata:
Algorithms
Applied sciences
Artificial intelligence
Artificial intelligence Evolutionary algorithms Non-linear programming Optimization
Computer science; control theory; systems
Connectionism. Neural networks
Evolutionary algorithms
Exact sciences and technology
Flows in networks. Combinatorial problems
Genetic algorithms
Genetic engineering
Genetic recombination
Methods
Non-linear programming
Nonlinear programming
Operational research and scientific management
Operational research. Management science
Optimization
Search theory
Studies
United States
Evolutionary algorithms
Non-linear programming
Artificial intelligence
Optimization
Genetic algorithm
Evolutionary algorithm
Optimal solution
Optimization method
Non linear programming
Global solution
Stochastic programming
ISSN:0305-0483, 1873-5274
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Over the past three decades Evolutionary Algorithms have emerged as a powerful mechanism for finding solutions to large and complex problems. A promising new evolutionary algorithm known as Differential Evolution (DE) was recently introduced and has garnered significant attention in the research literature. This paper introduces a modification to DE that enhances its rate of convergence without compromising solution quality. DE was recently shown to outperform several well-known stochastic optimization methods on an extensive set of test problems. Our Modified Differential Evolution (MDE) algorithm utilizes selection pressure to develop offspring that are more fit to survive than those generated from purely random operators. We demonstrate that MDE requires less computational effort to locate global optimal solutions to well-known test problems in the continuous domain.
Bibliografie:SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 14
ISSN:0305-0483
1873-5274
DOI:10.1016/j.omega.2004.04.009