Statistical Analysis Of Complex Network

Complex network analysis is inseparable from measurement and analysis of da-ta,so statistical methods play an important role in complex network analysis.This paper mainly studies from the most basic data visualization and network operation,to summarize the characteristics of the network,to the network data model,with abundant case illustrates how to use multiple R language to analyze all kinds of complex network expansion pack,here is the main use of igraph and ergm package,which also involves many packages such as network,sna,statnet,explains how to use igraph to create a network diagram.They provide different network data classes,data representation,storage,and graphical display.This paper studies the one can simultaneously comprehensive network topology and network node properties index of statistical model,random graph model(ERGM),the ERGM model and model constraints are studied and an application example is given based on the model.First,create an empty model,and as we gradually add predictors to the propaga-tion model,how to make AIC and BIC values smaller,which means that we are adding useful predictors for ERGM.Then the model is estimated by the MCMCM-LE method.Then the network is generated based on the model parameters and estimated parameters.Finally,the model diagnosis and simulation of the generated network are carried out.The main purpose of this study is to demonstrate the complex network anal-ysis techniques of various network packages and models in R.All R codes are also available.