Scholarlink

Federated learning (FL) is a decentralized machine learning (ML) paradigm designed to train models collaboratively across a large number of distributed devices or organizations while keeping the raw data localized [23]. Instead of pooling data into a central server, clients compute model updates locally and only share parameters or gradients with a coordinating server [6]. This framework preserves data privacy, reduces communication overhead, and allows leveraging heterogeneous and sensitive datasets that are otherwise difficult to aggregate.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Introduction to Federated Learning

  • 章节

  • 发表日期：

    2026年

查看全文
  • Kai Li,
  • Xin Yuan,
  • Wei Ni

摘要

Federated learning (FL) is a decentralized machine learning (ML) paradigm designed to train models collaboratively across a large number of distributed devices or organizations while keeping the raw data localized [23]. Instead of pooling data into a central server, clients compute model updates locally and only share parameters or gradients with a coordinating server [6]. This framework preserves data privacy, reduces communication overhead, and allows leveraging heterogeneous and sensitive datasets that are otherwise difficult to aggregate.