Scholarlink

<p>This paper introduces a fast-converging learning framework for residential energy consumption forecasting, named Federated Decision Tree (FEDT). FEDT integrates Federated Learning (FL) with Internet of Things (IoT) infrastructure and employs decision tree models for decentralized training across heterogeneous edge devices. The framework incorporates one client-side training strategy and four server-side aggregation mechanisms to collaboratively construct a global decision tree model. FEDT was fully implemented and evaluated on diverse platforms, including Raspberry Pi, Android smartphones, and personal computers (PCs). Experimental results show that FEDT stabilizes by the <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(17^{\text {th}}\)</EquationSource> </InlineEquation> round with approximately 80 trees per client, outperforming FedAVG, which requires 38 rounds to reach comparable performance. The fast-converging global training achieves forecasting errors below 2% across all clients. Additionally, a scalability analysis demonstrates that FEDT maintains efficient training and inference times, satisfactory server-side aggregation time and local client training time, and amount of communication traffic.</p>

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

Fast-converging federated decision trees for smart-home energy consumption prediction

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月15日

查看全文
  • Júlio Oliveira,
  • Yuri Santo,
  • André Riker,
  • Eirini Eleni Tsiropoulou,
  • Glaucio H. S. Carvalho

摘要

This paper introduces a fast-converging learning framework for residential energy consumption forecasting, named Federated Decision Tree (FEDT). FEDT integrates Federated Learning (FL) with Internet of Things (IoT) infrastructure and employs decision tree models for decentralized training across heterogeneous edge devices. The framework incorporates one client-side training strategy and four server-side aggregation mechanisms to collaboratively construct a global decision tree model. FEDT was fully implemented and evaluated on diverse platforms, including Raspberry Pi, Android smartphones, and personal computers (PCs). Experimental results show that FEDT stabilizes by the \(17^{\text {th}}\) round with approximately 80 trees per client, outperforming FedAVG, which requires 38 rounds to reach comparable performance. The fast-converging global training achieves forecasting errors below 2% across all clients. Additionally, a scalability analysis demonstrates that FEDT maintains efficient training and inference times, satisfactory server-side aggregation time and local client training time, and amount of communication traffic.