Each customer has specific purchase regularity. To predict the customer's purchase behavior, some researchers have built a static model which not considere the environmental change and the customer's characteristics. To dynamically predict customer purchase behavior, this paper introduces a posteriori estimation method. Based on the customer's purchased information, the method combine with the customer's current events, then applies the Bayes theorem to posteriori estimate the customer's purchase behavior. The new method not only predict individual customer's purchase behavior, but also improve the prediction accuracy. It will be helpful for the enterprises to optimize arrangements for the production and inventory and reduce operating costs.
展开▼
