基于密度峰值法的复杂网络聚类增长维度研究

Research on Clustering-growth Dimension of Complex Network Based on Density Peak Method

复杂网络是由个体或组织以及它们之间的关系所组成的结构。利用复杂网络的分形结构来解释和预测复杂网络的行为是目前的一个研究热点。分形维度是对复杂网络中分形结构的度量,为了更准确地对复杂网络分形结构进行度量,提出了一种基于密度峰值的方法来计算分形维度。该算法不同于之前选取一个种子节点或者将所有节点作为种子节点的方法,而是利用密度峰值法确定网络中的某几个核心节点作为种子,再计算网络的分形维度。仿真实验表明:基于密度峰值的方法比基于紧密度和原始方法得到更为精确的分形维度。

A complex network is a structure which is composed of individuals or organizations and the relationships among them.As a current research hotspot,the fractal structure of complex networks is used to explain and predict the behavior of complex networks.The fractal dimension is a measurement of the fractal structure in a complex network.In order to measure the fractal structure of a complex network more accurately,a method based on the density peak is proposed to calculate the fractal dimension.This algorithm is different from the previous method of selecting a seed node or using all nodes as seed nodes.Instead,it uses the density peak method to determine certain core nodes in the network as seeds,and then calculates the fractal dimension of the network.Simulation experiments show that the method based on density peaks can obtain more accurate fractal dimensions than the method based on compactness and original methods.