Prediction method and electronic equipment

By converting observable original features into latent features and building directed graphs for feature aggregation, the problem of inability to effectively consider potential factors in users' minds in the prior art is solved, and the accuracy and interpretability of predictions are improved.

CN119005395APending Publication Date: 2024-11-22LENOVO (BEIJING) LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411009079.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-25
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

When predicting problems, the prior art is difficult to fully represent potential factors in the user's mind, resulting in inaccurate prediction results, especially when hidden factors cannot be directly observed and their relationships are complex, the existing methods cannot effectively consider the impact of these factors.

Method used

By obtaining observable original features, performing principal component analysis and factor load weighting processing, unobservable potential features are generated, and potential features are used for prediction. At the same time, building a directed graph for feature aggregation, considering the topological structure between potential features. to improve prediction accuracy.

Benefits of technology

More accurate prediction results are achieved, which can effectively reflect the impact of important potential characteristics, and dynamically update the relationship between potential characteristics, improving the accuracy and interpretability of predictions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119005395A_ABST
    Figure CN119005395A_ABST
Patent Text Reader

Abstract

The embodiment of the invention provides a prediction method and electronic equipment. The prediction method comprises the following steps: acquiring a plurality of original features for a to-be-predicted problem; the original features are observable user features; analyzing and processing the plurality of original features to generate a plurality of potential features; the potential features are expressed by weighting a plurality of original features and corresponding factor loads, and the potential features are unobservable user features; and performing prediction according to the plurality of potential features to obtain a prediction result for the to-be-predicted problem. Therefore, the observable original features are converted into the unobservable potential features, and the potential features are used for prediction, so that the accuracy of the prediction result is improved.
Need to check novelty before this filing date? Find Prior Art