Improvisation of differential evolution for community detection

Most of the real world networks we encounter today are complex networks and one of the important characteristics of these networks is the community structure. Identifying communities in a complex network is classified as computably hard and thus many metaheuristic approaches have been proposed in the past. In this paper we propose an improved differential evolution based algorithm which exploits the structural similarity of the network to generate a better initial population leading to a more accurate identification of communities. We have tested our algorithm on various well-known real world and artificial networks.