Linear Spectral Statistics Based Hypothesis Test Of ER-Graph With Missing Data

Network analysis is a more and more important branches in modern science,which has been across several field of science.In recent years,complex network brought great advances for the analysis of big data,and random graph is the basic of complex network.In recent years,there are lots of models about how to tell a random graph is an ER graph,but have little effect when part of data is missing.Based on the study of conventional ER graph discrimination,this paper presents four methods of ER graph discrimination under the condition of missing data,and compares the advantages and disadvantages of these methods under various conditions.