Multi agent systems have been widely used to parallelizing processes in network-based computation environments. Parallelization of computational load could improve the performance of genetic algorithms and decrease their computational time. In this paper, a new distributed multi-agent system has been proposed for parallelizing genetic algorithm in usual networks. This dynamic approach could improve the performance of parallel genetic algorithm in a network-based context. We also provide a new migration policy for proposed architecture which leads us to attain a better performance for Parallel Genetic Algorithm. Experimental results show the efficiency of proposed method comparing to previous studies. © 2008 Springer-Verlag.
CITATION STYLE
Momeni, M., & Zamanifar, K. (2008). A dynamic multi agent-based approach to parallelizing genetic algorithm. In Communications in Computer and Information Science (Vol. 6 CCIS, pp. 830–833). https://doi.org/10.1007/978-3-540-89985-3_113
Mendeley helps you to discover research relevant for your work.