In order to meet the requirements for safety and latency in many IoT applications, intelligent decisions must be made right here right now at the network edge, calling for edge intelligence. To facilitate fast edge learning, this work advocates a platform-aided federated meta-learning architecture, where a set of edge nodes joint force to learn a meta-model (i.e., model initialization for adaptation in a new learning task) by exploiting the similarity among edge nodes as well as the cloud knowledge transfer. The federated meta-learning problem is cast as a regularized optimization problem, using Bregman Divergence between the edge model and the pre-trained model as the regularization. We then devise an inexact alternating direction method of multiplier (ADMM) based Hessian-free federated meta-learning algorithm, called ADMM-FedMeta, with inexact Hessian estimation. Further, we analyze the convergence properties and the rapid adaptation performance of ADMM-FedMeta for the general non-convex case. The theoretical results show that under mild conditions, ADMM-FedMeta converges to an $\epsilon$-approximate first-order stationary point after at most $\mathcal{O}(1/\epsilon^2)$ communication rounds. Extensive experimental studies on benchmark datasets demonstrate the effectiveness and efficiency of ADMM-FedMeta, and showcase that ADMM-FedMeta outperforms the existing baselines.

翻译：为了达到许多IOT应用的安全性和延缓性要求,必须现在就在网络边缘做出智能决定,要求获得精密情报。为了便利快速边缘学习,这项工作主张一个平台辅助的联邦化元学习架构,在这个架构中,一组边缘节点联合力量通过利用边缘节点之间的相似性以及云层知识转移来学习元模型(即适应新学习任务的模式初始化)。 联合的元学习问题被作为常规化优化问题,使用边际模型和预先培训的模型之间的分辨作为正规化。我们随后设计了一个基于Hesian-无Hesian-Federal-Fed-Metal 模型的不精确交替方向方法,称为ADMM-FedM-Feta, 以及不Exact Hesian 估计。此外,我们分析了ADMMM-F-FedMetta的趋同性特征和快速适应性表现,用于一般非convex案例。理论结果显示,在温度条件下,ADMMM-FM-MDMAS-MASMAS-MAS-MAS-MAS-ADASMAMAS ASMAD AS AS AS AS AS ASAL ASU ASU ASU ASUAL ASUAL ASU ASU ASU ASU ASU ASU ASU AS AS AS AS AS ASU AS AS AS AS AS AS AS AS AS ASU AS ASY ASY ASY ASU ASY ASY ASY ASY ASY AS AS AS ASU AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS ASU ASU AS AS ASU AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS AS