Federal learning model training method and device, electronic equipment and storage medium

Abstract

The invention provides a federal learning model training method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the sample alignment with a business side server; obtaining a plurality of generative models, and respectively obtaining current samples of the plurality of generative models; respectively acquiring alignment samples of the plurality of generative models from the current samples of the plurality of generative models; respectively inputting the alignment samples of the plurality of generative models into the corresponding generative models with initial generative model parameters to obtain a model score corresponding to each generative model, and sending the model score to a business party server; and receiving a model training score corresponding to each generative model sent by the business party server, and training the plurality of generative models according to the model training scores. Therefore, joint training between the business party server and the data provider server can be more efficient, meanwhile, communication burden is relieved, dependence on cryptology is not needed, and calculation complexity is reduced.

Description

technical field [0001] The present application relates to the technical field of data processing, and in particular to a training method, device, electronic device and storage medium of a federated learning model. Background technique [0002] Federated learning was first proposed by Google. The main idea is to build machine learning models based on data sets distributed on multiple devices while preventing data leakage. [0003] However, the currently popular federated learning technologies are all combined with cryptography to encrypt sensitive information that needs to be transmitted during the modeling process. Taking the more mature federated learning framework as an example, the framework provides a variety of federated machine learning Algorithms: decision tree, deep neural network, logistic regression, etc. The implementation of these algorithms relies on various methods of secure multi-party computing and cryptography, resulting in heavy communication burden, long m...

AI Technical Summary

Problems solved by technology

[0003] However, the currently popular federated learning technologies are all combined with cryptography to encrypt sensitive information that needs to be transmitted during the modeling process. Taking the more mature federated learning framework as an example, the framework provides a variety of federated machine learning Algorithms: decision tree, deep neural network, logistic regression, etc. The implementation of these algorithms relies on various methods of secure multi-party computing and cryptography, resulting in heavy communication burden, long modeling time, and strong cryptographic barriers. There are certain obstacles to the optimization of the algorithm

Images