为提高签名网络社区发现效果,解决其评估指标存在的数据耦合和依赖性,造成网络社区单指标优化存在较大局限性的问题,提出了基于位置修复和粒子置换的FSUD-PSO(fast sorting and uniform density of multi-objective particle swarm optimization)签名网络社区发现算法。首先,对签名网络模型进行研究,并在考虑数据耦合和依赖性前提下给出签名网络社区评价指标,以及多目标Pareto最优目标模型;其次,构建签名网络模型的多目标优化粒子编码与更新规则,并根据签名网络特点设计了位置修复和粒子置换策略,同时为提高多目标粒子群算法性能,设计了快速排序均匀密度的多目标粒子群算法FSUD-PSO;最后,通过标准测试集实验对比,验证了所提FSUD-PSO签名网络社区发现算法的有效性。%In order to improve the effect of signature network community discovery, and solve the evaluation indicator of the presence of data coupling and dependence, which leads some limitations of single index optimization in network community, this paper proposes signature network community discovery based on FSUD-PSO (fast sorting and uniform density of multi-objective particle swarm optimization) with location repair and particle replacement. Firstly, this paper studies the signature network model, and gives the community evaluation index of the signature network under the premise of considering the data coupling and dependence. Secondly, this paper builds a signature network model with particle coding and update rules for multi-objective optimization and network according to the characteristics of signature design repair and particle replacement, at the same time, in order to improve multi-objective particle swarm algorithm performance, it designs the FSUD-PSO algorithm. Finally, the effectiveness of the proposed FSUD-PSO signaturenetwork community is verified by comparing with the standard test sets.
展开▼