SOCCA: SOcial-based Colonial Competitive Algorithm

摘要

Evolutionary algorithms have been successfully applied as optimization tools in various applications. CCA-Colonial Competitive Algorithm - is a recently-developed socio-politically inspired evolutionary optimization algorithm. Initial population of CCA are divided in to some collections called empires and all the individuals in each empire move toward the best one to identify increasingly better area of the search space. This paper presents SOCCA (SOcial-based Colonial Competitive Algorithm) as an improved version of CCA. In each empire of our algorithm, a social network is generated, assigning neighbours for each individual to interact with and individuals move toward their best neighbours. To assess the SOCCA capability, it was applied to four benchmark optimization functions and results show that the SOCCA algorithm has the ability of finding the global minimum. Also, we compared results with CCA, which indicates SOCCA superiority. Our findings show that SOCCA may provide better performance for some other applications.