首页|Researchers from Henan Normal University Provide Details of New Studies and Find ings in the Area of Intelligent Systems (A Lightweight and Personalized Edge Fed erated Learning Model)
Researchers from Henan Normal University Provide Details of New Studies and Find ings in the Area of Intelligent Systems (A Lightweight and Personalized Edge Fed erated Learning Model)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Research findings on Machine Learning - Intelligent Systems are discussed in a new report. According to news originati ng from Xinxiang, People's Republic of China, by NewsRx correspondents, research stated, "As a new distributed machine learning paradigm, federated learning has gained increasing attention in the industry and research community. However, fe derated learning is challenging to implement on edge devices with limited resour ces and heterogeneous data." Financial support for this research came from National Natural Science Foundatio n of China (NSFC). Our news journalists obtained a quote from the research from Henan Normal Univer sity, "This study aims to realize a lightweight and personalized model through p runing and masking with insufficient resources and heterogeneous data. Particula rly, the server first downloads the subnetwork to the client according to the ma sk, and client prunes the subnetwork with the alternating direction method of mu ltipliers (ADMM), so as to remove the unimportant parameters and reduce the cost of training and communication. At the same time, mask is used to mark the pruni ng condition of the model. Then, the unpruned parts and masks of local models ar e transmitted to the server for aggregation. The experimental results showed tha t the accuracy of the proposed model was improved by 9.36%, and the communication cost was reduced by 1.45 times compared with state-of-the-art mod els."
XinxiangPeople's Republic of ChinaAs iaIntelligent SystemsMachine LearningHenan Normal University
NETL
NSTL
