Improved Ant Colony Clustering Algorithm In Customer Classification Research And Application

With the Internet's wide application,the rapid growth of e-commerce platforms like taobao and jd.com,which has led to the fierce market competition,traditional physical retail enterprises facing a huge number of consumers,using the original popular marketing mode is not only costly but also has no advantages.The classification of customers plays an important role in the relationship management of the physical retail industry.It can make accurate marketing of different user groups,improve the efficiency of enterprise marketing strategy and obtain better marketing effect.Traditional data analysis method is difficult to deal with massive user data,so the clustering analysis algorithm based on data mining is used to realize the clustering of enterprise customers.Firstly,the paper introduces the related technologies of data mining and clustering analysis,including the concept of data mining,the process and methods of data mining and the definition and evaluation criteria of cluster analysis technology,and enumerates several common clustering algorithms.Then briefly introduced some related concepts of physical retail customer classification,and explained the importance and necessity of entity retail customer classification.The paper focuses on the standard ant colony clustering algorithm and its development and existing problems,and tries to improve ant clustering algorithm from several aspects:the convergence speed of the algorithm,the movement direction selection in the clustering process and the dependence of the algorithm on parameter setting.Then we propose an improved ant cluster algorithm,and we use the modified ant cluster algorithm to cluster all the data in the Iris dataset,and the results show that the improved algorithm has a higher rate of accuracy,and the rate of convergence of the algorithm is faster,and it proves that the improved ant cluster algorithm is effective.Finally,the improved clustering algorithm is applied to the customer classification of an entity retail enterprise,according to the total consumption amount,total consumption number and recent consumption time,according to the user's behavior data,the user is gathered to different clusters,and the clustering results are obtained,and the characteristics of each category are analyzed according to the results,and the customers of each cluster are analyzed,and the targeted marketing and decision-making plan is put forward.