参照零模型的双层网络结构相关性检测

发布时间：2018-09-14 20:10

【摘要】：多层网络是近年来提出的新型复杂网络模型,在级联故障分析、信息传播、链路预测和网络同步等诸多领域均有广泛应用。多层网络的任意两层网络结构间往往存在关联性和耦合性,如何检测两层网络之间是否具有结构相关性并对其进行定量刻画是一个非常重要且亟待解决的问题。本文首先分3个层次总结并提出度量双层网络结构相关性的方法与统计量,其中第1层次是检测双层网络整体上的连接相关性,第2层次检测两层网络所有节点之间整体上的度度相关特性,第3层次是检测双层网络富节点之间的连接相关性。由于这3种相关性的计算和度量都依赖于网络统计量进行,而统计量的绝对数值往往没有意义,因此提出了以多种双层网络的零模型作为参照物,通过假设检验方法来量化双层网络之间的结构相关性,分析了这种结构相关性存在的内在机理。最后,使用一个实证双层网络——全球语言多层网络验证了本文研究范式的有效性。本文研究可检测出实证多层网络中任意两层之间复杂的耦合作用。
[Abstract]:Multilayer network is a new complex network model proposed in recent years. It is widely used in cascade fault analysis, information transmission, link prediction and network synchronization. There is always correlation and coupling between any two layers of multilayer networks. How to detect the structural correlation between the two layers and describe them quantitatively is a very important and urgent problem to be solved. In this paper, the method and statistics of measuring the correlation of bilevel network structure are summarized and put forward in three levels. The first level is to detect the connection correlation of the bilevel network as a whole. The second level detects the overall degree correlation between all nodes in the two-layer network, and the third level detects the connection correlation between the rich nodes in the two-layer network. Since the calculation and measurement of these three correlations depend on the network statistics, and the absolute values of the statistics are often meaningless, the zero model of a variety of double-layer networks is proposed as the reference. The structural correlation between bilevel networks is quantified by hypothesis testing, and the inherent mechanism of the structural correlation is analyzed. Finally, the effectiveness of the research paradigm is verified by using an empirical two-tier network, the Global language Multilayer Network (GLN). In this paper, the complex coupling between any two layers in an empirical multilayer network can be detected.
【作者单位】：大连民族大学信息与通信工程学院;贵州大学贵州省公共大数据重点实验室;
【基金】：国家自然科学基金项目(61374170,61603073) 大连市青年科技之星培育计划项目(2015R091) 中央高校基本科研业务费专项资金项目(DC201502060201)
【分类号】：O157.5

【相似文献】