Federated learning (FL) is an emerging paradigm for training deep neural networks (DNNs) in distributed manners. Current FL approaches all suffer from high communication overhead and information leakage. In this work, we present a federated learning algorithm based on evolution strategies (FedES), a zeroth-order training method. Instead of transmitting model parameters, FedES only communicates loss values, and thus has very low communication overhead. Moreover, a third party is unable to estimate gradients without knowing the pre-shared seed, which protects data privacy. Experimental results demonstrate FedES can achieve the above benefits while keeping convergence performance the same as that with back propagation methods.

翻译：联邦学习（FL）是一种以分布式方式训练深度神经网络（DNNs）的新兴范式。当前的联邦学习方法普遍存在通信开销高和信息泄露的问题。本文提出一种基于进化策略的联邦学习算法（FedES），这是一种零阶训练方法。与传输模型参数不同，FedES仅通信损失值，因此具有极低的通信开销。此外，由于无法获知预共享的随机种子，第三方无法估计梯度，从而保护了数据隐私。实验结果表明，FedES在保持与反向传播方法相同收敛性能的同时，能够实现上述优势。