federated learning (FL) is an emerging distributed machine learning method
that empowers in-situ model training on decentralized edge devices. However,
multiple simultaneous training activities could overload res
提出一种用于协调和训练多个同时进行的联邦学习任务的系统,名为 MAS(Merge and Split),通过多任务架构将多个任务合并为一个任务进行训练,并根据任务之间的亲和度在训练过程中将其拆分成两个或更多任务继续训练,实验证明 MAS 方法在性能优化方面胜过其他方法,同时减少训练时间 2 倍并降低 40% 的能源消耗。