Artificial Bee Colony Algorithm And Its Application In Clustering Algorithm

With the rapid development of Internet and information technology in the world,there is a growing demand for network resources sharing.These shared data information has caused data explosion and information explosion.How to find a scientific and reasonable way to help people from a large number of complex and disorderly data to filter out effective and reliable information is an urgent need to study the problem.Data mining is an effective method to solve this problem,which can help people make correct and effective decision after analyzing and processing specific data.Clustering algorithm is widely used in the field of data mining,so it has become a research hotspot of many scholars and experts.In recent years,many researchers have applied the swarm intelligence optimization algorithm to the clustering analysis field,which has improved the clustering effect and stability to a certain extent.Artificial bee colony algorithm,as a new group intelligent optimization algorithm,is developing very fast,but it is still in the initial stage,and the algorithm model is not perfect enough,which attracts many researchers to improve the basic artificial bee colony algorithm.On the basis of classical clustering method,this paper analyzes the limitations of clustering algorithm,and then studies the theoretical knowledge of artificial bee colony algorithm,and improves the artificial bee colony algorithm,then combines the improved artificial bee colony algorithm to Traditional K-means clustering algorithm.The main work is as follows:(1)In view of the deficiency of the basic artificial bee colony algorithm,this paper proposes an improved artificial bee colony algorithm.Social learning plays a key role in group intelligence.In basic artificial bee colony algorithm,social learning is realized mainly by leading bees and following bees' searching behavior.However,in the search strategy of basic artificial bee colony algorithm,the search range is determined randomly,and the algorithm cannot control the search range.This will lead to the problem that the search precision is not high and the convergence rate of the algorithm is slow.In this paper,we first introduce the central solution idea with more optimal solutions to improve the search efficiency and speed up the convergence of the colony.Then,considering the effect of the artificial bee colony algorithm on the global search in different evolutionary periods And the requirement of local mining ability are different.On the basis of the central solution,the dynamic adjustment factor is added to the honey source search strategy,which makes the algorithm automatically adjust the search range in different evolutionary periods and enhance the algorithm's The search ability and the mining ability of the algorithm are well balanced.The experimental results show that the convergence rate and the accuracy of the improved algorithm are greatly improved.(2)K-means clustering algorithm is commonly used in data mining algorithms,because its principle is easy to understand,easy to implement the advantages,has been widely used.However,K-means algorithm has a strong dependency on the choice of the initial cluster center and is easy to fall into the local optimal solution in the process of optimizing the data set.In this paper,the improved artificial bee colony algorithm is applied to In the clustering problem,the K-means algorithm is optimized.The experimental results show that the improved algorithm reduces the dependence on the initial clustering center to a certain extent,has strong stability,and the clustering effect is obviously improved.