CN-122001777-A - Information determination method, device, terminal, network equipment and medium

Abstract

The application discloses an information determining method, a device, a terminal, network side equipment and a medium, which belong to the technical field of communication; the first device determines performance information of the VFL model based on first intermediate data, wherein the first intermediate data is intermediate data corresponding to verification data, and the verification data is used for verifying the VFL model.

Inventors

• CHENG SIHAN
• CHONG WEIWEI

Assignees

• 维沃移动通信有限公司

Dates

Publication Date

20260508

Application Date

20241107

Claims (20)

1. 1. An information determining method, the method comprising: The first device receives first intermediate data from the second device; The first device determining performance information of a longitudinal federal learning VFL model based on the first intermediate data; The first intermediate data is intermediate data corresponding to verification data, and the verification data is used for verifying the VFL model.
2. 2. The method of claim 1, wherein the first device determining performance information of a VFL model based on the first intermediate data comprises: the first device determining output data of the VFL model based on the first intermediate data; the first device determines the performance information based on the output data and tag data, the tag data corresponding to the verification data.
3. 3. The method according to claim 1 or 2, characterized in that the method further comprises: the first device sends first information to the second device; Wherein the first information includes at least one of: The verification data; Identification information of the verification data; a data range of the verification data; The first indication information is used for indicating that the first information comprises related information of the verification data; And second indication information for indicating the verified related information.
4. 4. The method of claim 3, wherein prior to the first device transmitting the first information to the second device, the method further comprises: The first device determining to validate the VFL model based on the second information; Wherein the second information includes at least one of a number of training iteration rounds of the VFL model, a loss function of the VFL model.
5. 5. The method according to any one of claims 1 to 4, further comprising: the first device sends third information to the second device; wherein the third information includes at least one of: Training data; identification information of training data; A data range of training data; The third indication information is used for indicating that the third information comprises relevant information of training data; fourth indication information, the fourth indication information is used for indicating relevant information of training; wherein the training data is used to train the VFL model.
6. 6. The method according to any one of claims 1 to 5, further comprising: the first device determines training data and validation data based on the aligned sample data, the sample data being related to the VFL model.
7. 7. The method according to any one of claims 1 to 6, wherein the verification data comprises at least one of: Verifying data used in the stage; Actual usage data; Model reasoning data; model test data.
8. 8. The method of any one of claims 1 to 7, wherein the validation data comprises one or more sets of data.
9. 9. An information determining method, the method comprising: the second device sends first intermediate data to the first device; The first intermediate data is used for determining performance information of the VFL model, and the first intermediate data is intermediate data corresponding to verification data, wherein the verification data is used for verifying the VFL model.
10. 10. The method according to claim 9, wherein the method further comprises: The second device receives first information from the first device; The second device determining the first intermediate data based on the first information; Wherein the first information includes at least one of: The verification data; Identification information of the verification data; a data range of the verification data; The first indication information is used for indicating that the first information comprises related information of the verification data; And second indication information for indicating the verified related information.
11. 11. The method according to claim 9 or 10, characterized in that the method further comprises: the second device receives third information from the first device; The second device determining second intermediate data based on the third information; the second intermediate data is intermediate data corresponding to training data, the training data is used for training the VFL model, and the third information includes at least one of the following: The training data; identification information of the training data; a data range of the training data; The third indication information is used for indicating that the third information comprises relevant information of training data; Fourth indication information, the fourth indication information is used for indicating relevant information of training.
12. 12. The method according to any one of claims 9 to 11, wherein the verification data comprises at least one of: Verifying data used in the stage; Actual usage data; Model reasoning data; model test data.
13. 13. The method according to any one of claims 9 to 12, wherein the verification data comprises one or more sets of data.
14. 14. An information determining apparatus, characterized in that the apparatus comprises a first receiving module and a first processing module; The first receiving module is used for receiving first intermediate data from the second equipment; the first processing module is used for determining performance information of the VFL model based on the first intermediate data received by the first receiving module; The first intermediate data is intermediate data corresponding to verification data, and the verification data is used for verifying the VFL model.
15. 15. The apparatus of claim 14, wherein the first processing module is configured to determine output data of the VFL model based on the first intermediate data, and to determine the performance information based on the output data and tag data, the tag data corresponding to the verification data.
16. 16. The apparatus according to claim 14 or 15, further comprising a first transmission module; the first sending module is used for sending first information to the second equipment; Wherein the first information includes at least one of: The verification data; Identification information of the verification data; a data range of the verification data; The first indication information is used for indicating that the first information comprises related information of the verification data; And second indication information for indicating the verified related information.
17. 17. The apparatus of claim 16, wherein the first processing module is further configured to determine to verify the VFL model based on second information before the first sending module sends the first information to the second device; Wherein the second information includes at least one of a number of training iteration rounds of the VFL model, a loss function of the VFL model.
18. 18. The apparatus according to any one of claims 14 to 17, further comprising a first transmission module; the first sending module is used for sending third information to the second equipment; wherein the third information includes at least one of: Training data; identification information of training data; A data range of training data; The third indication information is used for indicating that the third information comprises relevant information of training data; fourth indication information, the fourth indication information is used for indicating relevant information of training; wherein the training data is used to train the VFL model.
19. 19. The apparatus of any of claims 14 to 18, wherein the first processing module is further configured to determine training data and validation data based on the aligned sample data, the sample data being related to the VFL model.
20. 20. The apparatus according to any one of claims 14 to 19, wherein the authentication data comprises at least one of: Verifying data used in the stage; Actual usage data; Model reasoning data; model test data.

Description

Information determination method, device, terminal, network equipment and medium Technical Field The application belongs to the technical field of communication, and particularly relates to an information determining method, an information determining device, a terminal, network side equipment and a medium. Background At present, in the model training process, the model periodically calculates performance information of the model, such as accuracy, namely the percentage of the correct number of model predictions and the total number of predictions, specifically, whether the training is correct or not can be judged by comparing a predicted value generated by the model with a label of the training, so that the performance information of the model is calculated. The model training performance can be objectively evaluated through the performance information of the model, and further model parameters are continuously adjusted to improve the training effect of the model. Longitudinal federal learning (VERTICAL FEDERATED LEARNING, VFL) is a model training mode, and is essentially a combination of features, and is suitable for scenes with more overlapping users and less overlapping features. By combining the different data characteristics of the common samples of the participants, the VFL can increase the feature dimensions of the training samples, thereby training to a better model. However, for the VFL model, there is no solution to how the VFL server determines the performance information of the VFL model, and thus the model training performance cannot be objectively evaluated, resulting in poor training effect of the VFL model. Disclosure of Invention The embodiment of the application provides an information determining method, an information determining device, a terminal, network side equipment and a medium, which can determine the performance information of a VFL model, further objectively evaluate the model training performance and improve the training effect of the VFL model. In a first aspect, an information determining method is provided and performed by a first device, the method including the first device receiving first intermediate data from a second device, the first device determining performance information of a VFL model based on the first intermediate data, wherein the first intermediate data is intermediate data corresponding to verification data used to verify the VFL model. In a second aspect, an information determining method is provided and performed by a second device, where the method includes the second device sending first intermediate data to a first device, where the first intermediate data is used to determine performance information of a VFL model, the first intermediate data is intermediate data corresponding to verification data, and the verification data is used to verify the VFL model. In a third aspect, an information determining apparatus is provided, where the apparatus includes a first receiving module and a first processing module, the first receiving module is configured to receive first intermediate data from a second device, and the first processing module is configured to determine performance information of a VFL model based on the first intermediate data received by the first receiving module, where the first intermediate data is intermediate data corresponding to verification data, and the verification data is used to verify the VFL model. In a fourth aspect, an information determining apparatus is provided, where the apparatus includes a second sending module, and a second sending module configured to send first intermediate data to a first device, where the first intermediate data is used to determine performance information of a VFL model, and the first intermediate data is intermediate data corresponding to verification data, where the verification data is used to verify the VFL model. In a fifth aspect, there is provided an information determining apparatus configured to perform the steps of the method as described in the first aspect or to implement the steps of the method as described in the second aspect. In a sixth aspect, there is provided a terminal comprising a processor and a memory storing a program or instructions executable on the processor, which when executed by the processor, performs the steps of the method according to the first aspect, or performs the steps of the method according to the second aspect. In a seventh aspect, a terminal is provided, which includes a processor and a communication interface, where the communication interface is configured to receive first intermediate data from a second device, and the processor is configured to determine performance information of a VFL model based on the first intermediate data received by the first receiving module, where the first intermediate data is intermediate data corresponding to verification data, and the verification data is used to verify the VFL model. Or the communication interface is