Customer loss prediction method, device and equipment and computer readable storage medium

Abstract

The embodiment of the invention discloses a customer loss prediction method, a customer loss prediction device, customer loss prediction equipment and a computer readable storage medium. The method comprises the following steps: processing historical log data of customer information of a to-be-predicted project according to specific time granularity to obtain time sequence characteristic data; inputting the time sequence characteristic data into a pre-constructed tree model to obtain tree model characteristics; and inputting the time sequence characteristic data and the tree model characteristics into a customer loss prediction model, the customer loss prediction model being composed of a long and short term memory network model and a convolutional neural network model and comprising a hybrid model in cross-layer connection, and the output of the model being each customer loss probability of the to-be-predicted project. According to the method, the defects existing in manual feature extraction in related technologies are overcome, the customer loss prediction accuracy is improved, and the method has good universality.

Description

technical field [0001] The embodiments of the present invention relate to the technical field of machine learning, and in particular, relate to a method, device, equipment, and computer-readable storage medium for predicting customer churn. Background technique [0002] Customer churn is a phenomenon in which customers give up continuing to use a certain service provided by an enterprise, such as the phenomenon that mobile phone users in the telecommunications industry leave the network, and members of Internet paid subscription services do not renew their subscriptions. With the increasingly fierce business competition in the market, customer loss is becoming more and more likely to happen. [0003] The loss of customers can have a big impact on the company's profits, and the cost of acquiring new customers is much higher than retaining existing customers. Therefore, in order to pursue sustainable development, the company must do a good job of retaining existing customers....

AI Technical Summary

Problems solved by technology

[0006] The traditional customer churn prediction method based on artificial features combined with classic classifiers requires a lot of manpower to extract features manually, and it is not universal. In a different business scenario, the method of extracting features or the effect of features may become invalid, which leads to end customers. The accuracy of churn prediction depends on the quality of artificial features, and it does not have the versatility of cross-business prediction

Images