Research And Application Of Important Node Identification Method Based On Multi-layer Aging Network

In recent years,research on complex networks has been endless,and the research level of complex networks is gradually deepened.At present,the research on multilayer temporal networks is the focus of current research,and the research on the importance of node recognition as a complex network has practical application significance.This paper mainly focuses on the research and application of node importance identification in multilayer temporal network,mainly as follows:(1)For the two-layer time network,this paper proposes an improved directed weighted projection technology based on the existing weighted projection technology.This technology considers the direction of the connected edges of the nodes,the weight ratio and the asymmetry of the weights,and constructs a weight distribution adjacency matrix,and defines the importance value of the nodes according to the weight distribution adjacency matrix,thereby achieving the node importance of a class of nodes.Effective means of sexual identification.Compared with the existing algorithms through experiments,it is found that the accuracy and time balance of the new algorithm are better reflected.(2)For the multilayer temporal network,this paper constructs a mathematical model of the multi-layer aging network and a sixth-order tensor representation,taking into account the connectivity of the nodes on the multi-layer and time layers,and introduces the PageRank centrality The idea is to use a third-order tensor to represent the PageRank iterative centrality algorithm.In experimental verification,by comparing with the PageRank centrality algorithm using a fifth-order tensor and the PageRank centrality algorithm of a multilayer network using a fourth-order tensor,a new algorithm and a centrality algorithm using a fifth-order tensor are obtained.There is a linear relationship,and the algorithm is more robust than the other two algorithms.(3)In view of the multilayer temporal network node importance of applications,applied in this paper,from two aspects,first,this section put forward based on the second floor time network node importance recognition of link prediction algorithm research,through the study of the application of the improved weighted projection technology,considering the weight,the topology on the edge,and the importance of nodes,we can find predict the link performance is better than traditional algorithm.Second,based on important node of the prescription of multilayer temporal network nodes similarity identification research and application of the recommended algorithm used to build the mathematical model and tensor of multilayer temporal,said the use of node similarity index,is applied to the feature vector in the iteration algorithm of centricity,and to evaluate important node,which is applied to recommend.