Prediction-based federated learning communication optimization method and system

Abstract

The invention relates to the field of federated machine learning, and discloses a prediction-based federated learning communication optimization method and system. The method comprises the steps thatfirstly, a global model and global variables needed in the method are initialized, each terminal user carries out local model training according to local data of the terminal user, and local model updating is obtained; then, a cloud center predicts the local model update of each terminal user according to the historical model update trend of each terminal user; then, a prediction error threshold value of each terminal user is set by calculating changes of prediction updating and global model loss functions adopted by the terminal user, and the prediction error threshold value comprises two steps of setting an initial threshold value and setting a dynamic threshold value; finally, a global model updating strategy is designed according to the set prediction error threshold value, and the cloud center adopts accurate prediction updating to replace local model updating to calculate global model updating. The problem of high communication cost caused by frequent transmission of update parameters between the terminal users and the cloud center in the federated learning technology is solved.

Description

technical field [0001] The present invention relates to the field of federated machine learning, and more specifically, to a prediction-based federated learning communication optimization method and system, which are used to solve the high communication cost caused by frequent transmission of update parameters between terminal users / equipment and cloud centers in federated learning technology question. Background technique [0002] As an important branch of artificial intelligence, machine learning has been successfully and widely used in various fields such as pattern recognition, data mining and computer vision. Due to the limited computing resources of terminal devices, the current training of machine learning models usually adopts a cloud-based method. In this method, the data collected by terminal devices, such as pictures, videos, or personal location information, must be uploaded to the cloud center Centralize the training of the model. However, uploading the user's...

AI Technical Summary

Problems solved by technology

[0009] In order to more effectively solve the high communication cost problem of federated learning, the present invention proposes a prediction-based federated learning communication optimization method and system

Images