Predicting Customer Behaviour Involving Risk Factor using Bivariate Hierarchical Bayesian Approach

摘要：

Current approach to predict changes in individual customer behaviour ignored the character of randomicity, heterogeneity and irrelevance between purchase interval and money. Bivariate Hierarchical Bayesian(BHB) approach supposes prior distribution of customer behavior as Bivariate Logarithmic Normal distribution, and deduces posterior distribution using Hierarchical Bayesian Theory and computes parameter with the method of Markov chain Monte Carlo, which considers the character of customer behavior synthetically, and can predict customer behaviour involving interpurchase time, monetary value and customer risk.. The model is applied to medical instruments sale data to predict customer changes and shows more precisely than traditional models. Based on distribution of customer behavior, the concept of customer risk is brought out, including churn risk, declining risk and fluctuating risk, which can be calculated using probability density curve. This value prediction considers risk can be used managerially as a signal for the firm to use some type of intervention to keep customers.

关键词： Bivariate hierarchical Bayesian approach Logarithmic normal distribution Markov chain Monte Carlo Customer risk Customer behavior

作者: WANG Jie WANG Haiwei

作者单位: School of Management, Harbin Institute of Technology, Harbin, P.R.China 150001

会议类型: 国际会议

会议名称: The 5th International Conference on Product Innovation Management(第五届产品创新管理国际会议 ICPIM 2010)

会议地点: 武汉

会议语种:英文

页码: 974-981

在线出版日期: 2010-07-10（万方平台首次上网日期，不代表论文的发表时间）