With the increasing popularity of e-commerce and the rapid development. The competition among e-commerce companies becomes fiercer (Zhang, 2015). As known, the most significant matter is to retain customers by providing the best services along with a suitable price. Therefore, with the rapid increase of e-commerce transaction volume and intense competition in the market to meet the high demand from customers, it is necessary to attract customers through customised services and targeted strategies to increase customer loyalty. On the other hand, e -commerce customer churn shows nonlinear changes and asymmetrical customer categories. E-commerce customer churn data has a typical sample imbalance which means that the number of churn samples is significantly larger than the number of non-churned samples or vice versa. This research proposes the methods or models imposed on e-commerce to proactively reduce customer churn.