Customer loss prediction method and device

Abstract

The invention provides a customer loss prediction method and apparatus. The method comprises the steps of obtaining to-be-predicted customer data; and inputting the to-be-predicted customer data intoa customer loss prediction model to calculate data of lost customers, the customer loss prediction model being obtained by training historical customer data of a plurality of customers. By means of the mode, the customer loss prediction method and system can accurately predict the customer loss and find the customer possibly losing, so that the customer is saved in time, the customer saving cost is saved, the saving effect is improved, and the customer loss problem is effectively solved.

Description

technical field [0001] The invention relates to the technical field of data processing, in particular to a method and device for predicting customer loss. Background technique [0002] Due to the huge impact of customer churn on corporate profits, the current problem of customer churn has received extensive attention. In the increasingly competitive banking industry, the problem of customer churn also needs to be solved urgently. The existing customer churn prediction method is to monitor the total assets of customers. When the total assets of customers drop significantly in a short period of time, for example, by 250,000, it is determined that the customer may be lost and the customer is retained. With this customer churn prediction method, the prediction is not accurate enough, the cost of customer retention is high and the benefits are small. Contents of the invention [0003] In view of this, the present invention provides a method and device for predicting customer ...

AI Technical Summary

Problems solved by technology

[0003] In view of this, the present invention provides a method and device for predicting customer churn, which are used to solve the problem of inaccurate prediction of existing customer churn prediction methods

Images