蚁群聚类优化算法在零售客户分类中的应用

Application of ant colony clustering optimizationalgorithm in retail customer classification

随着网上购物热潮的到来,企业拥有的客户数据激增。挖掘并分析出隐藏在客户数据中的信息,实现客户群进行划分,对提高企业盈利有显著作用。鉴于此,研究从移动策略、观察半径、概率转换函数等三个方面进行蚁群聚类算法的优化,并以蚁群聚类优化算法实现客户数据的聚类分析。研究结果显示,与标准蚁群聚类算法相比,蚁群聚类优化算法的平均错误个数降至9.6个,平均运行时间降至31.23s;客户群1对应的零售客户消费总次数均值、消费总金额均值均最高,依次为19次、36439元。这些结果说明蚁群聚类优化算法能够实现零售客户的分类,且分类耗时短、分类质量高。

with the advent of online shopping boom,the customer data owned by enterprises has increased dramatically.Mining and analyzing the information hidden in the customer data and realizing the division of customer groups have a significant effect on improving enterprise profitability.In view of this,this paper studies the optimization of ant colony clustering algorithm from three aspects of mobile strategy,observation radius and probability transfer function,and uses ant colony clustering optimization algorithm to realize the clustering analysis of customer data.The results show that,compared with the standard ant colony clustering algorithm,the average number of errors of the ant colony optimization algorithm is reduced to 9.6,and the average running time is reduced to 31.23 s;the average number of times and total consumption amount of retail customers corresponding to customer group 1 are the highest,which are 19 times and 36439 yuan respectively.These results show that the ant colony clustering optimization algorithm can achieve the classification of retail customers,and the classification time is short and the classification quality is high.