On the Efficiency of Crossover Operators in Genetic Algorithms with Binary Representation (CROSBI ID 564176)
Prilog sa skupa u zborniku | izvorni znanstveni rad | međunarodna recenzija
Podaci o odgovornosti
Picek, Stjepan ; Golub, Marin
engleski
On the Efficiency of Crossover Operators in Genetic Algorithms with Binary Representation
Genetic Algorithm (GA) represents robust, adaptive method successfully applied to various optimization problems. To evaluate the performance of the genetic algorithm, it is common to use some kind of test functions. However, the ”no free lunch”’ theorem states it is not possible to find the perfect, universal solver algorithm. To evaluate the algorithm, it is necessary to characterize the type of problems for which that algorithm is suitable. That would allow conclusions about the performance of the algorithm based on the class of a problem. In performance of a genetic algorithm, crossover operator has an invaluable role. To better understand performance of a genetic algorithm in a whole, it is necessary to understand the role of the crossover operator. The purpose of this paper is to compare larger set of crossover operators on the same test problems and evaluate their’s efficiency. Results presented here confirm that uniform and two-point crossover operators give the best results but also show some interesting comparisons between less used crossover operators like segmented or half-uniform crossover.
Evolutionary computation ; Genetic algorithms ; Crossover operator ; Efficiency ; Binary representation ; Test functions
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
Podaci o prilogu
167-172.
2010.
objavljeno
Podaci o matičnoj publikaciji
roceedings of the 2th WSEAS International Conference on Evolutionary Computing, EC'10
Munteanu, Viorel ; Raducanu, Razvan ; Dutica, Gheorghe ; Balas, Valentina Emilia ; Gavrilut, Alina
Iași: WSEAS Press
978-960-474-195-3
1790-5109
Podaci o skupu
WSEAS International Conference on Evolutionary Computing, EC'10
predavanje
13.06.2010-15.06.2010
Iaşi, Rumunjska