基于粒子群优化的复杂网络社区挖掘

Complex Network Community Mining Based on Particle Swarm Optimization

为解决复杂网络社区结构挖掘的优化问题,根据复杂网络拓扑结构的先验知识,提出一种基于离散粒子群优化的社区结构挖掘算法。将粒子的位置和速度定义在离散环境下,设计粒子的更新规则,在不需要事先指定社区个数的前提下自动判断网络的最佳社区个数,给出局部搜索算子,该算子可以帮助算法跳出局部最优解,提高算法的收敛速度和全局寻优能力。实验结果表明,与iMeme-net算法相比,该算法能够准确地挖掘出复杂网络中隐藏的社区结构,且执行速度较快。

In order to solve the problem of community mining optimization from complex network,according to the prior knowledge of the topology structure of complex network,a complex network community mining algorithm based on Particle Swarm Optimization（PSO）is proposed. In the proposed algorithm,particle＇s position and velocity are redefined in discrete case,particle＇s update principles is redesigned,the proposed algorithm can automatically determine the best community numbers without knowing it in advance. In order to improve the global search ability of the proposed algorithm,a local search operator is designed,and this operator can help the algorithm to jump out of local optimum and improves the convergence speed. Experimental results demonstrate that the proposed algorithm can efficiently dig out the community structures hidden behind complex networks,and the execution speed is much faster than that of i Meme-net algorithm.