A federated learning model training method and system

Abstract

The invention discloses a federated learning model training method and system, wherein the method includes: a central server sends initial parameters to each node device; each node device performs model training based on the initial parameters and a local data set, and obtains a gradient value; the central server Generate a key pair and send the public key to each node device; each node device uses the public key to encrypt the gradient value to obtain an encrypted gradient; each node device calculates the encrypted gradient sum in a point-to-point manner, and sends the encrypted gradient sum to to the central server; the central server decrypts the encrypted gradient sum through the private key, obtains the gradient sum, and sends it to each node device; the node device updates the parameters of the model based on the gradient sum, according to the updated parameters and the local data set. In the next round of model training, until the training stop condition is met, this method can guarantee the privacy data of all participants and the accuracy of model training.

Description

technical field [0001] The present application relates to the field of information technology, in particular to a federated learning model training method and system. Background technique [0002] Federated learning is dedicated to solving the problem of multi-user collaborative model training without disclosing their respective data sets. For example, in the medical big data modeling scenario, each hospital has different patient sample data. Due to the limited amount of data owned by each hospital, if the model is trained only based on its own data, the effect of the model is difficult to meet expectations due to the limitation of the sample size. If the data of various companies can be aggregated for training, the accuracy of the model can be greatly improved. However, due to the competition among hospitals and the requirements for the privacy protection of patient data, it is impossible for hospitals to release their own customer data to any third party, and it is not f...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the problem that the existing federated learning method is difficult to balance the accuracy and privacy, the present invention provides a federated learning model training method and system

Images