Title :
Quasi-parallel genetic algorithms with different communication topologies
Author :
Wang, Guan ; Wu, Degang ; Szeto, K.Y.
Author_Institution :
Dept. of Phys., Hong Kong Univ. of Sci. & Technol., Hong Kong, China
Abstract :
Systematic testing of parallel genetic algorithm defined on simple topologies such as ring, band, and fully connect graph are made with two benchmark problems, the 0-1 Knapsack Problem and the Weierstrass function. With fixed communication rule between sub-populations, our numerical results indicate that the band configuration is superior to other topologies tested. A heuristic understanding in terms of the compromise between exploitation and exploration in the solution space is suggested.
Keywords :
genetic algorithms; graph theory; knapsack problems; parallel algorithms; Knapsack Problem; Weierstrass function; band configuration; communication topologies; quasiparallel genetic algorithms; Biological cells; Exchange rates; Genetic algorithms; Network topology; Parallel processing; Scalability; Topology; Knapsack Problem; Optimization; Parallel Genetic Algorithms; Topology; Weierstrass Function;
Conference_Titel :
Evolutionary Computation (CEC), 2011 IEEE Congress on
Conference_Location :
New Orleans, LA
Print_ISBN :
978-1-4244-7834-7
DOI :
10.1109/CEC.2011.5949690