Prediction system, method and device for user repayment timeliness and electronic equipment

Abstract

The invention discloses a prediction method for user repayment timeliness. The method comprises the steps that a first client acquires first identification data and a first feature data set of a target user; a second prediction feature data set of the first feature data set at a second client is predicted by using the dual model, and the first identification data and encrypted data of the second prediction feature data set are sent to the second client; a first encryption weighted value of the first feature data set is obtained and the first encryption weighted value is sent to the server; thesecond client determines a second target feature data set of the target user at the second client by using the received encrypted data, obtains a second encryption weighted value of the second targetfeature data set and sends the second encryption weighted value to the server; the server sums the received first encryption weighted value and second encryption weighted value, decrypts the summation result to obtain target label data representing the repayment timeliness of the target user, and sends the target label data to the first client; wherein the first client is one of a sales enterprise terminal or a bank terminal, and the second client is the other of the sales enterprise terminal or the bank terminal.

Description

technical field [0001] The invention belongs to the field of finance, and in particular relates to a prediction system, method, device and electronic equipment about the timeliness of user repayment. Background technique [0002] In the financial field, based on business needs, banks often need to predict whether users can repay loans such as credit cards in a timely manner, but the bank's own user data may not be sufficient for prediction. At the same time, sales companies also expect to predict whether users can repay the loan in time to carry out user portraits. Therefore, banks and sales companies expect to jointly complete the prediction process of user repayment timeliness without revealing the private data of their respective users. [0003] In this case, the concept of "federated learning" came into being. Federated learning is actually an encrypted distributed machine learning technology. Sales companies and banks can use the longitudinal federated learning method...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem of insufficient user data sets shared by sales companies and banks in vertical federated learning, and to improve the predictive effect of the longitudinal federated learning models of the two on the timeliness of user repayment, the embodiment of the present invention provides a user repayment timeliness A predictive system, method, device, electronic device, storage medium, and a federated learning model training method

Images