Scholarlink

The advent of Federated Learning (FL) has brought about a revolutionary change in the field of machine learning, enabling the decentralised training of models across a multitude of devices while simultaneously maintaining the confidentiality of the data. In contrast to conventional centralized methodologies, FL maintains the localisation of data, with only model updates being shared. This methodology enhances model generalisation and stability without compromising data sovereignty. A variety of machine learning techniques, including support vector machines (SVMs) and decision trees, can be effectively utilised within the context of FL frameworks. SVMs offer efficient solutions for classification tasks with minimal computational overhead, while decision trees provide interpretable models for both classification and regression. This paper explores the application of these methods in FL settings, highlighting their advantages and potential use cases in diverse industries, particularly in manufacturing. Furthermore, it discusses the integration of reinforcement learning with FL, emphasising its potential for enhancing intelligent and adaptable decentralised systems.

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

Multifaceted Applications of Federated Learning: Beyond Neural Networks

  • 会议论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年

查看全文
  • Tatjana Legler,
  • Vinit Hegiste,
  • Martin Ruskowski

摘要

The advent of Federated Learning (FL) has brought about a revolutionary change in the field of machine learning, enabling the decentralised training of models across a multitude of devices while simultaneously maintaining the confidentiality of the data. In contrast to conventional centralized methodologies, FL maintains the localisation of data, with only model updates being shared. This methodology enhances model generalisation and stability without compromising data sovereignty. A variety of machine learning techniques, including support vector machines (SVMs) and decision trees, can be effectively utilised within the context of FL frameworks. SVMs offer efficient solutions for classification tasks with minimal computational overhead, while decision trees provide interpretable models for both classification and regression. This paper explores the application of these methods in FL settings, highlighting their advantages and potential use cases in diverse industries, particularly in manufacturing. Furthermore, it discusses the integration of reinforcement learning with FL, emphasising its potential for enhancing intelligent and adaptable decentralised systems.