Training method, device, electronic device and medium for user satisfaction prediction model

By acquiring different types of sample data and processing features, combined with gradient enhancement algorithm, the training effect and generalization ability of the user satisfaction prediction model are improved, and the poor training effect and generalization effect of the user satisfaction prediction model in the existing technology are solved. Poor problem.

CN117057459BActive Publication Date: 2024-12-13CHINA MOBILE INTERNET CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310946533.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-07-28
Publication Date
2024-12-13
Estimated Expiration
2043-07-28

AI Technical Summary

Technical Problem

The training effect of the existing user satisfaction prediction model is poor and the generalization effect is poor.

Method used

By obtaining different types of sample data (subjective dimension and objective dimension data), the user's satisfaction preference category is determined, and the sample characteristics are processed according to the category, the initial satisfaction prediction model is trained, and the gradient enhancement algorithm and multiple superpositions are constructed. Integrate the model to improve the generalization ability of the model.

Benefits of technology

It effectively improves the training effect and generalization ability of the user satisfaction prediction model, and solves the problems of poor training effect and poor generalization effect of the user satisfaction prediction model in the prior art.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117057459B_ABST
    Figure CN117057459B_ABST
Patent Text Reader

Abstract

The present disclosure proposes a training method, device, electronic device and medium for a user satisfaction prediction model, including: determining to obtain first sample data and second sample data of a user to be predicted, and determining the satisfaction preference category of the user to be predicted based on the first sample data and the second sample data, and then determining a first sample feature of the first sample data and a second sample feature of the second sample data, and processing the first sample feature according to the satisfaction preference category to obtain a first target sample feature, and processing the second sample feature according to the satisfaction preference category to obtain a second target sample feature, training an initial satisfaction prediction model based on the first target sample feature and the second target sample feature to obtain a user satisfaction prediction model, and training the user satisfaction prediction model by combining multi-dimensional data sample data can improve the training effect of the user satisfaction prediction model, thereby improving the generalization ability of the user satisfaction prediction model.
Need to check novelty before this filing date? Find Prior Art