Fair guarantee-oriented federated learning model optimization method and system

Abstract

The invention provides a fairness guarantee-oriented federated learning model optimization method and system. The method comprises the steps of receiving first optimization parameters of a to-be-optimized classification model and to-be-tested images sent by a plurality of clients; calculating the accuracy of the to-be-optimized classification model of each client on different to-be-tested images, and obtaining a contribution degree matrix formed by the accuracy corresponding to all clients; obtaining a total accuracy rate corresponding to each client according to the contribution degree matrix, and constructing an optimization model according to the total accuracy rate corresponding to each client, the first number, the second number and a variance between the total accuracy rates corresponding to all clients; and according to the optimal solution of the optimization model, allocating a second optimization parameter to each client, so that each client optimizes the to-be-optimized classification model according to the second optimization parameter. According to the method, the performance of the to-be-optimized classification model after all the clients are optimized is comprehensively improved, and a distributed system formed by the server and the clients can be stably and continuously used for a long time.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a model optimization method and system for federated learning oriented to fairness guarantee. Background technique [0002] Federated learning is a distributed learning framework that can effectively help multiple institutions to perform data usage and machine learning modeling while meeting the requirements of user privacy protection, data security, and government regulations, and thus has attracted widespread attention. [0003] Federated learning is a multi-party collaborative system. The motivation for each participant to participate is to get a better model than local training. In order to ensure that the system can last for a long time and stably, it is necessary to further ensure the fairness of the system . Therefore, in the federated learning scenario, how to achieve fairness is an urgent problem to be solved. [0004] In order to solve the above problems, there are ...

AI Technical Summary

Problems solved by technology

[0007] The present invention provides a model optimization method and system for federated learning oriented to fairness guarantee, which is used to solve the problem that in the prior art, federated learning only considers opportunity fairness or result fairness, and cannot make all clients obtain model parameters with good performance , and then it is difficult to ensure that the system can continue for a long time and stably, so as to achieve accurate optimization of model parameters for all clients

Images