Ada-FFL:Adaptive computing fairness federated learning

作     者：Yue Cong Jing Qiu Kun Zhang Zhongyang Fang Chengliang Gao Shen Su Zhihong Tian 

作者机构：The Cyberspace Institute of Advanced TechnologyGuangzhou UniversityGuangzhouChina Pengcheng LabShenzhenChina Carnegie Mellon UniversityPittsburghPennsylvaniaUSA Mohamed Bin Zayed University of Artificial IntelligenceAbu DhabiUnited Arab Emirates 

出 版 物：《CAAI Transactions on Intelligence Technology》 (智能技术学报（英文）)

年 卷 期：2024年第9卷第3期

页      面：573-584页

核心收录：

学科分类：070801[理学-固体地球物理学] 07[理学] 08[工学] 0708[理学-地球物理学] 0816[工学-测绘科学与技术] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：National Natural Science Foundation of China,Grant/Award Number:62272114 Joint Research Fund of Guangzhou and University,Grant/Award Number:202201020380 Guangdong Higher Education Innovation Group,Grant/Award Number:2020KCXTD007 Pearl River Scholars Funding Program of Guangdong Universities(2019) National Key R&D Program of China,Grant/Award Number:2022ZD0119602 Major Key Project of PCL,Grant/Award Number:PCL2022A03 

主　　题：adaptive fariness aggregation fairness federated learning non-IID 

摘      要：As the scale of federated learning expands,solving the Non-IID data problem of federated learning has become a key challenge of *** existing solutions generally aim to solve the overall performance improvement of all clients;however,the overall performance improvement often sacrifices the performance of certain clients,such as clients with less *** fairness may greatly reduce the willingness of some clients to participate in federated *** order to solve the above problem,the authors propose Ada-FFL,an adaptive fairness federated aggregation learning algorithm,which can dynamically adjust the fairness coefficient according to the update of the local models,ensuring the convergence performance of the global model and the fairness between federated learning *** integrating coarse-grained and fine-grained equity solutions,the authors evaluate the deviation of local models by considering both global equity and individual equity,then the weight ratio will be dynamically allocated for each client based on the evaluated deviation value,which can ensure that the update differences of local models are fully considered in each round of ***,by combining a regularisation term to limit the local model update to be closer to the global model,the sensitivity of the model to input perturbations can be reduced,and the generalisation ability of the global model can be *** numerous experiments on several federal data sets,the authors show that our method has more advantages in convergence effect and fairness than the existing baselines.