Optimization method for gain model and time sequence prediction model

By training a gain model and introducing an auxiliary loss function to optimize ranking capabilities, the problem of inaccurate ranking by the gain model is solved, and accurate assessment and ranking of individual response differences are achieved.

CN120911531APending Publication Date: 2025-11-07ZHEJIANG UNIV +1
6 Cites 0 Cited by

Patent Information

Application Number
CN202510987686.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-16
Publication Date
2025-11-07

AI Technical Summary

Technical Problem

Existing gain models cannot accurately assess the differences in individual responses to interventions, resulting in insufficient ranking capabilities. Traditional assessment methods are biased, leading to incorrect rankings.

Method used

By acquiring a dataset, training a gain model, introducing an auxiliary loss function to optimize the model, ensuring that the conditional average causal effect of persuasive individuals is greater than that of unpersuasive individuals, and redefining the principle gain curve to ensure that persuasive individuals are ranked ahead of unpersuasive individuals.

Benefits of technology

It significantly improves the ranking accuracy of the gain model, avoids the biases brought about by traditional evaluation methods, and enhances the model's ability to accurately assess individual response differences.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses an optimization method for a gain model. The optimization method comprises the following steps: acquiring a data set for training the gain model; wherein a data sample in the data set comprises a covariable, a binary processing variable and a binary result variable; inputting a data sample in the data set into the gain model, so that the gain model performs training calculation corresponding to a task target of the gain model for the data sample; wherein the training calculation comprises the steps of estimating a conditional average causal effect of preset processing performed on the individuals relative to the preset result based on the data sample, and sorting the individuals in the data set according to a sequence of values of the estimated conditional average causal effect from large to small; calculating model loss generated by training the gain model; and adjusting model parameters of the gain model based on the calculated model loss so as to complete training of the gain model.
Need to check novelty before this filing date? Find Prior Art