Masuda Naoki, Miwa Hiroyoshi, Konno Norio
Laboratory for Mathematical Neuroscience, RIKEN Brain Science Institute, 2-1, Hirosawa, Wako, Saitama, 351-0198, Japan.
Phys Rev E Stat Nonlin Soft Matter Phys. 2004 Sep;70(3 Pt 2):036124. doi: 10.1103/PhysRevE.70.036124. Epub 2004 Sep 30.
Many real networks are complex and have power-law vertex degree distribution, short diameter, and high clustering. We analyze the network model based on thresholding of the summed vertex weights, which belongs to the class of networks proposed by Phys. Rev. Lett. 89, 258702 (2002)]. Power-law degree distributions, particularly with the dynamically stable scaling exponent 2, realistic clustering, and short path lengths are produced for many types of weight distributions. Thresholding mechanisms can underlie a family of real complex networks that is characterized by cooperativeness and the baseline scaling exponent 2. It contrasts with the class of growth models with preferential attachment, which is marked by competitiveness and baseline scaling exponent 3.
许多真实网络都很复杂，具有幂律顶点度分布、短直径和高聚类性。我们分析了基于顶点权重总和阈值化的网络模型，该模型属于《物理评论快报》89, 258702 (2002)所提出的网络类别。对于许多类型的权重分布，都会产生幂律度分布，特别是具有动态稳定缩放指数2、现实的聚类性和短路径长度。阈值化机制可能是一类以合作性和基线缩放指数2为特征的真实复杂网络的基础。它与以竞争性和基线缩放指数3为特征的带偏好依附的增长模型类别形成对比。
