Memetic clonal selection algorithm with EDA vaccination for unconstrained binary quadratic programming problems

► MCSA-EDA proposes an estimation of distribution algorithm (EDA) vaccination, as hypermutation and recombination operator, to effectively use the global statistical information to guide the search. ► MCSA-EDA introduces three components, i.e., EDA vaccination, fitness uniform selection scheme and a...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Expert systems with applications Ročník 38; číslo 6; s. 7817 - 7827
Hlavní autoři: Cai, Yiqiao, Wang, Jiahai, Yin, Jian, Zhou, Yalan
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier Ltd 01.06.2011
Témata:
Algorithms
Antibodies
Computer simulation
Estimation of distribution
Expert systems
Fitness
Fitness uniform selection
Memetic clonal selection algorithm
Operators
Performance enhancement
Quadratic programming
Searching
Tabu search
Unconstrained binary quadratic programming problem
Fitness uniform selection
Tabu search
Estimation of distribution
Memetic clonal selection algorithm
Unconstrained binary quadratic programming problem
ISSN:0957-4174, 1873-6793
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:► MCSA-EDA proposes an estimation of distribution algorithm (EDA) vaccination, as hypermutation and recombination operator, to effectively use the global statistical information to guide the search. ► MCSA-EDA introduces three components, i.e., EDA vaccination, fitness uniform selection scheme and adaptive Tabu search, to overcome the deficiencies of the conventional clonal selection algorithm. ► MCSA-EDA has been shown as a competitive approach for unconstrained binary quadratic programming problems (UBQP) and thus as a powerful and effective framework for combinatorial optimization. This paper presents a memetic clonal selection algorithm (MCSA) with estimation of distribution algorithm (EDA) vaccination, named MCSA-EDA, for the unconstrained binary quadratic programming problem (UBQP). In order to improve the performance of the conventional clonal selection algorithm (CSA), three components are adopted in MCSA-EDA. First, to compensate for the absence of recombination among different antibodies, an EDA vaccination is designed and incorporated into CSA. Second, to keep the diversity of the population, a fitness uniform selection scheme (FUSS) is adopted as a selection operator. Third, to enhance the exploitation ability of CSA, an adaptive tabu search (TS) with feedback mechanism is introduced. Thus, MCSA-EDA can overcome the deficiencies of CSA and further search better solutions. MCSA-EDA is tested on a series of UBQP with size up to 7000 variables. Simulation results show that MCSA-EDA is effective for improving the performance of the conventional CSA and is better than or at least competitive with other existing metaheuristic algorithms.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ObjectType-Article-2
ObjectType-Feature-1
ISSN:0957-4174
1873-6793
DOI:10.1016/j.eswa.2010.12.124