自组织映射神经网络与VIP客户识别研究

Recognizing YIPs based on self-organizing map neural network

根据著名的“帕雷托80／20原理”,20％的客户(VIP客户)贡献了公司80％的利润。因此,准确地识别VIP客户对确保企业的收益是至关重要的。本文试图运用自组织映射神经网络聚类分析方法从海量的客户数据中准确地识别VIP客户,并针对VIP客户的消费行为特点提出了一系列相应的营销策略,以规避企业丢失VIP客户的风险、增加VIP客户的忠诚度和使得企业收益最大化。本论文采用自组织映射神经网络聚类识别VIP客户是一种非常有意义的尝试和创新。

This paper mainly expatiates on the principle and application of SOM (i. e. Self-Organizing Map) Neural Networks clustering algorithm in recognizing VIPs from a huge database of customers on the Internet. Through SOM Neural Networks clustering algorithm, VIPs can be clustered and recognized objectively and scientifically, and this will help enterprises avoid the risk of taking the uniform-service strategy for all the customers or taking the subjective and unscientific rank-service strategy to lost VIP customers. According to the Pareto 80/20 Principle: 20% of all the customers (namely VIPs) for an enterprise contribute 80% of all its profit, scientifically recognizing VIPs based on SOM Neural Network is a very significant and comparatively novel method. In this paper, the author hopes eagerly that her study on recognizing VIPs can help enterprises increase the loyalty of VIPs and maximize the profit of enterprises.