Technology Research On Community Detection In Complex Networks Based On Cuckoo Search Algorithm

In real life,many complicated systems,such as mobile communication systems,electric power systems,interpersonal social systems,and intercity transportation systems all have their own complex network properties.The individuals that make up these systems are called nodes,the connection between individuals is called the edge of a complex network.Nodes converge with each other or separate from each other bring about the cluster,We call it community,In order to analyze a complex network,find the unobtrusive communities in the network and analyze the community structure is necessary,Due to the great market value of the science and application of complex networks,community detection technology has been discussed by most researchers.Based on the cuckoo algorithm,this paper designs two intelligent optimization methods for the community detection problem of objective function optimization,An algorithm based on cuckoo search algorithm combining gene retention and discrete difference strategy(GDCSA)and multi-objective community detection algorithm combining hill-climbing search strategy(MCD-HSS).Depart from a single objective,GDCSA used modularity as the objective function.In order to overcome the shortcomings of traditional cuckoo algorithms commonly used in continuous function optimization problems,random walk combined with high-quality gene retention is used to update the bird's nest position in global search.This method can increase the flexibility of the update process and improve the search efficiency,In the elite search stage,by using an improved logistic function to map the random steps in Levy flight,At the same time,in order to ensure the diversity of the population,the differential evolution method was adopted.Experiments show that the algorithm has the accuracy of division,the evaluation index on some networks is better than the comparison algorithm.Starting from the multi-objective optimization direction,the MCD-HSS algorithm uses two improved functions as the objective function.The global search is consistent with the single-objective search strategy.The local search introduces a simple and easy-to-implement hill climbing algorithm to further improve the nest quality,Experiments show that in the hierarchical structure,the algorithm can obtain different structural divisions,and some evaluation indicators are better than the comparison algorithm.