Federated Learning (FL) is a recent development in distributed machine learning that collaboratively trains models without training data leaving client devices, preserving data privacy. In real-world FL, the training set is distributed over clients in a highly non-Independent and Identically Distributed (non-IID) fashion, harming model convergence speed and final performance. To address this challenge, we propose a novel, generalised approach for incorporating adaptive optimisation into FL with the Federated Global Biased Optimiser (FedGBO) algorithm. FedGBO accelerates FL by employing a set of global biased optimiser values during training, reducing 'client-drift' from non-IID data whilst benefiting from adaptive optimisation. We show that in FedGBO, updates to the global model can be reformulated as centralised training using biased gradients and optimiser updates, and apply this framework to prove FedGBO's convergence on nonconvex objectives when using the momentum-SGD (SGDm) optimiser. We also conduct extensive experiments using 4 FL benchmark datasets (CIFAR100, Sent140, FEMNIST, Shakespeare) and 3 popular optimisers (SGDm, RMSProp, Adam) to compare FedGBO against six state-of-the-art FL algorithms. The results demonstrate that FedGBO displays superior or competitive performance across the datasets whilst having low data-upload and computational costs, and provide practical insights into the trade-offs associated with different adaptive-FL algorithms and optimisers.
翻译:联邦学习联盟(FL)是分布式机器学习的最新发展,它通过合作培训模型,而没有培训数据离开客户设备,从而保护数据隐私。在现实世界FL,培训组合以高度非独立和同样分布(非IID)的方式在客户中分布,损害了模式趋同速度和最后性能。为了应对这一挑战,我们提出了一个创新的、通用的方法,将适应性优化纳入FL,与FedGBO(FedGBO)全球双向优化算法相结合。 FedGBBO(FedGBO)的算法加快了FL。我们还利用一套全球偏差的选调派值加快了FL,在培训中将非IID数据的“客户驱动”从非客户中减少,同时受益于适应性优化。 我们显示,在FedGBO, 更新全球模式可以作为集中化培训,使用偏差的梯度和软件更新,并运用这一框架证明FDGBO在使用动力-SG(SG) 低交易(SG) 进行大规模实验,使用4 FFL基准数据设置数据(CIFFFFA-RO-RO-SAL-SL) 3SL(CI-ROD-SL) 和SIM-SIM-SL) 和SIM-SIM-SIM-SAL-SL-SIM-SB-SB-SB-S-S-SB-SB-SB-SB-SB-SB-SB-SB-SL-SL-SL-SD-SD-SD-SD-SD-SD-SL 和SL-SD-SD-SD-SD-SD-SL-SL-SL-SL-SL-SL-SL-SL-SL-SL-SD-SB-SB-SD-SD-SL-SD-SB-SB-SB-SB-SB-SB-SB-SB-SL-SB-SL-SL-SB-SB-SL-SL-SL-S-S-S-S-S-SL-SB-SB-S