A Novel Ant Colony Optimization Algorithm with Dynamic Control Population for Community Detecting

Abstract

Nowadays, complex networks have driven great interests of scholars. As a special characteristic of a network, the community structure has wide research prospects. Many current algorithms are adopted for detecting the potential community structure, in which the ant colony algorithm is a typical one. However, the computational cost of the ant colony is too high which limits its performance. In this paper, we propose a novel ant colony optimization algorithm with dynamic control population. In the proposed algorithm, when a certain condition is reached, the number of ants starts to decrease based on the proposed rules. The efficiency of the proposed algorithm is estimated through comparing with the classical ant colony algorithm in real-world networks. Experiments show that the proposed algorithm has apparently lower computational cost, while the quality of the division is reserved relatively.