融入客户价值特征和情感特征的网络客户流失预测研究

Study on the Value Feature and the Emotion Feature to Predict the Web Customer Churn

在网络客户流失预测中,引入网络客户价值特征和情感特征,构建基于客户行为特征分析的网络客户流失预测模型结构。以国内某电子商务网站客户流失预测为实例,利用支持向量机与人工神经网络、逻辑斯蒂回归、朴素贝叶斯、决策树等方法进行对比研究,发现网络客户流失预测模型结构能在考虑预测模型准确性的条件下,兼顾预测模型的可解释性,是研究网络客户流失问题的有效方法。

A model structure on the analysis of customer behavior was constructed by using customer value feature and emotion feature for web customer churn prediction.The structure was used in web customer churn prediction for home electronic business carrier by comparing with support vector machine,artificial neural network,logistic regression,naive Bayesian classifier and decision tree.It is found that the structure has better accuracy with acceptable explanation of prediction model,and provides an effective approach for studying on two types of errors for web customer churn.