基于主成分与神经网络的我国电商企业客户流失风险预警研究

发布时间：2018-01-26 13:22

本文关键词： 电商企业客户流失风险预警主成分分析神经网络　出处：《青岛科技大学》2017年硕士论文　论文类型：学位论文

【摘要】：在我国电商企业迅猛发展的形势下,企业逐渐将竞争的焦点转移到客户资源的争夺上。伴随着电商企业对客户资源争夺竞争激烈程度的加大,客户越来越处于一种不稳定的状态,导致各电商企业由于频繁且大量的客户流失延长了企业利润的回收周期。在这种各大电商面临客户流失问题的大数据形势下,电商企业需要将关注的重点聚焦在如何根据客户已经体现出的个体特性及发生的行为规律预测企业未来可能发生的流失情况,并结合这些规律制定出行之有效的应对措施,以使企业适应竞争环境的变化、更加接近企业利益最大化的目标,这同时也是电子商务信息管理领域的一个重要课题。为了根据客户个体和行为规律预测企业未来的客户流失,十分关键且首要的一步是了解影响企业客户流失的风险影响因素。针对这点,论文在理论研究的基础上结合访谈的方法得到可能影响电子商务企业客户流失的重要因素,构成客户流失风险预警指标体系,并以调查问卷的方式对目标电商平台使用人群进行调查,获取研究所需样本数据。在问卷调查之后的研究过程中,应用科学的方法对调查问卷的信度和效度进行检验。随后应用主成分分析方法提取出与电商企业客户流失关系最大最有代表性的指标,作为客户流失风险预警模型的输入层,根据聚类分析得到的四个不同客户流失风险预警等级作为模型输出层,构建了电商企业客户流失风险预警模型。最后,在检验了模型的正确性之后,在文章研究基础上提出电商企业应对客户流失风险不同警情的防范和控制建议。
[Abstract]:With the rapid development of e-commerce enterprises in China, enterprises gradually shift the focus of competition to the competition of customer resources. Customers are increasingly in an unstable state. As a result of frequent and a large number of customer turnover, e-commerce enterprises have extended the profit recovery cycle. Under this situation, e-commerce companies are faced with customer churn problem in big data situation. Ecommerce enterprises need to focus on how to predict the loss of enterprises in the future according to the individual characteristics and behavior rules of customers. In order to adapt to the change of the competitive environment, and to make the enterprises more close to the goal of maximizing the interests of enterprises, the effective countermeasures are worked out in combination with these laws. This is also an important subject in the field of e-commerce information management. In order to predict the future customer turnover according to the individual and behavior rules of customers. A key and first step is to understand the risk factors that affect customer turnover. On the basis of theoretical research, this paper combines the method of interview to get the important factors that may affect the customer turnover of e-commerce enterprises, and constitutes the early warning index system of customer churn risk. And by the way of questionnaire to the target ecommerce platform users of the survey to obtain the research needs of the sample data. In the research process after the questionnaire. The reliability and validity of the questionnaire were tested by using scientific methods. Then principal component analysis was used to extract the largest and most representative indicators of customer turnover in e-commerce enterprises. As the input layer of the customer churn risk early warning model, according to the four different customer loss risk warning levels obtained by cluster analysis as the model output layer, an e-commerce enterprise customer loss risk warning model is constructed. Finally. After checking the correctness of the model, this paper puts forward some suggestions on how to prevent and control the different situations of customer churn risk in e-commerce enterprises on the basis of the research in this paper.
【学位授予单位】：青岛科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F724.6;F274

【参考文献】