Community detection in complex networks based on immune genetic algorithm

Abstract

As many of the community detection methods based on intelligent optimization algorithms suffer from degeneracy, unsatisfactory optimization ability, complex computational process, requiring priori knowledge, etc., a community detection method in complex networks based on immune Genetic Algorithm( GA) was proposed. The algorithm combined the improved character encoding with the corresponding genetic operator, and automatically acquired the optimal community number and the community detection solution without the priori knowledge. Immune principle was introduced into selection operation of GA, which maintained the diversity of individuals, and therefore improved the intrinsic degeneracy of GA. By utilizing the local information of the network topology structure in initialization population, crossover operation and mutation operation, the search space was compressed and the optimization ability was improved. The simulation results on both computer-generated networks and real-world networks show that the algorithm acquires the optimal community number and the community detection solution, and has a higher accuracy. This indicates the algorithm is feasible and valid for community detection in complex networks.