The Improved Estimation of Distribution Algorithms for Community Detection

Abstract

Based on the analysis of local monotonic of modularity function, this paper designs a fast and effective mutation operator, and then proposes an improved Estimation of Distribution Algorithm (EDA) for solving community detection problem. The proposed algorithm is tested on basic network and big scale complex network. Experimental results show that this algorithm can get 0.419 8 for the average Q function while running 100 times, has better performance than Girvan-Newman(GN) algorithm, Fast Newman (FN) algorithm and Tasgin Genetic Algorithm (TGA).