Causal contribution node analysis system applied to horizontal federated learning

Abstract

This invention discloses a causal contribution node analysis system applied to horizontal federated learning. This system combines model performance evaluation with causal inference mechanisms, enabling fair and traceable assessment of the contributions of data provided by multiple participants to the final model without sharing original data. The system primarily comprises three modules: a dataset confirmation module, a contribution rule confirmation module, and a revenue-sharing suggestion module. Further, it includes several modules covering data selection, data governance, propensity score modeling, influence function analysis, Shapley value calculation, and contribution weighting. The system can identify the marginal contribution of rare or strategic data and consider the data governance costs and computational inputs of participating parties to generate final revenue-sharing suggestions. This invention promotes multi-party collaboration in federated learning environments, ensures data rights, and is beneficial for application in automated revenue-sharing systems such as smart contracts.

Claims

1. A causal contribution node analysis system applied to horizontal federated learning, comprising: A dataset verification module group is used to process and verify the dataset of federated learning. The dataset verification module group includes: a dataset allocation and selection module, which is used to divide the dataset according to the data purpose set by the model initiator, and can set the initial profit sharing weight according to the data characteristics of the participants. And a data governance module, which is used to verify and process whether the data provided by participants meets the format, structure and consistency requirements of federated learning; a group of contribution confirmation rules, which is used to establish and reveal the contribution evaluation rules of federated learning, including: a contribution influencing factor selection module, which is used to set cost items and weighting factors to adjust data contributions; and a causal feature selection module, which is used to determine the features required for bias score modeling through statistical analysis, domain knowledge or feature interaction selection. The system includes a contribution rule transparency module to compile and reveal the settings for participants to confirm; a group of profit-sharing suggestion modules to calculate and provide profit-sharing suggestions for each participant, which includes: a model performance contribution module to analyze the marginal contribution of a single data point to the model's effectiveness; a causal feature contribution calculation module to evaluate the causal contribution of data based on propensity scores; and a contribution quantification module to integrate model performance, causal analysis, and weighting factors to calculate the contribution ratio of each participant. And a profit-sharing suggestion module, which is used to generate profit-sharing suggestions for each participant.

2. The system as described in claim 1, wherein, The dataset allocation and selection module is further used to classify data according to different data usage types.

3. The system as described in claim 1, wherein, The contribution factor selection module is further used to set at least one of the following cost factors: raw data assetization cost, additional human resource cost, computing resource cost, or other quantifiable cost.

4. The system as described in claim 1, wherein, The contribution influence factor selection module is further used to set at least one of the following weighting mechanisms: model initiator data weighting, missing value data weight reduction, or special group data weighting.

5. The system as described in claim 1, wherein, The model performance contribution module further includes: a baseline model generator, which generates a model without specific participant data; and an incremental contribution analyzer, which calculates the performance improvement of the model after adding specific participant data, and can output Shapley values ​​or influence functions as contribution evaluation metrics.

6. The system as described in claim 1, wherein, The causal feature contribution calculation module uses logistic regression to model propensity scores.

7. The system as described in claim 1, wherein, The contribution quantification module is used to sum up the contribution values ​​of individual data points and calculate the contribution ratio of each participant's overall data to the final model.

8. The system as described in claim 1, wherein, The profit-sharing suggestion module is used to generate profit-sharing suggestions based on the model's profit results and the proportion of participants' contributions, and can support the import of the suggestions into smart contracts for automated profit-sharing.

9. The system as described in claim 1, wherein, This system can be applied to multi-party data analysis scenarios in medical, financial, or other fields where privacy needs to be protected, and supports the quantification of data value in highly heterogeneous data situations.

Images