A Hybrid CS/DE Algorithm for Global Optimization

Mansooreh Soleimani, Shahriar Lotfi, and Amirhossein Ghodrati


Cuckoo Search, Global optimization, Differential Evolution, Metaheuristic, Hybrid evolutionary algorithm


This paper presents the hybrid approach of two nature inspired metaheuristic algorithms; Cuckoo Search (CS) and Differential Evolution (DE). In this paper we improve on CS algorithm by the use of DE algorithm operators. In standard CS algorithm, each cuckoo bird lays only one egg, however in nature is not the same and lay more eggs. Thus, we used the DE algorithms’ strategy (Crossover) for their reproduction. This operator increases the diversity of solutions. Furthermore, we added mutation operator in order to reduce the chance of eggs to be discovered, because cuckoo birds are specialized in mimicry in color and pattern of the host birds. This operator prevents the algorithm from being trapped into local optimum. Experimental results are examined with some standard benchmark functions in different dimensions and the obtained results show a promising performance of this algorithm.

Important Links:

Go Back