Gossip Learning as a Decentralized Alternative to Federated Learning

前言

最近开始看一些关于边缘计算的论文了，怕自己忘性大，所以用博客来记录这些论文的重要内容和自己的理解

Abstract

• Federated Learning is distributed machine learning approach for computing models over data collected by edge devices. The data is not collected centrally, but a master-worker architecture is applied where a master node performs aggregation and edge devices are the workers.
• Gossip Learning also assumes that the data remains at the edge devices, but it requires no aggregation server or any central component.
• Surprisingly, gossip learning actually outperforms federated learning in all the scenarios where the training data are distributed uniformly over the nodes, and it performs comparably to federated learning overall.

Introduction

• 相比起于参数服务器模式，federated learning’s framework is optimized so as to minimize communication between the server and the wo