Genetic algorithm optimizing modularity for community detection in complex networks

Community structure of complex network by genetic algorithm has drawn much attention of researchers in various field in recent years. Genetic algorithm is easy to trap into local optimal and also easy to obtain unstable solutions. In order to solve this problem, an effective mutation method combined with node-to-community membership function that is based on the local information of each node in networks is proposed. The heuristic initialization algorithm is also used to produce initial population with the accurate and diverse, which can further improve the search efficiency and the stability of the algorithm. In addition, we used modularity as fitness function, which can simplify the algorithm, in this paper the method has been tested on computer-generated network and real world network, and compared to classical algorithm. The experimental results have been shown that the proposed algorithm has a higher precision and effective for community structure.