What question did this study set out to answer?

The study aims to advance federated learning methodologies by tackling challenges posed by dual-skewed non-IID data.

March 28, 2026Open Access

Supplementary Material for "FedDCG: A Federated Learning Algorithm Based on Dynamic Consensus Graph"

YXYi XuAnhui University ZLZhouliang LuAnhui University XLXiao LiuDeakin University

Key Points

The study aims to advance federated learning methodologies by tackling challenges posed by dual-skewed non-IID data.
Presents FedDCG, a new federated learning algorithm.
Analyzes the effects of dual-skewed non-IID data on federated learning performance.
Incorporates consensus-based weighting and sample-size-aware aggregation techniques.
FedDCG improves aggregation efficiency in federated learning.
Demonstrates effective handling of imbalance in non-IID data.
Enhances model performance and convergence rate over traditional methods.

Abstract

This record contains the supplementary material for the paper "FedDCG: A Federated Learning Algorithm Based on Dynamic Consensus Graph". The supplementary file provides additional technical details that support the main manuscript. The associated paper studies federated learning under dual-skewed non-IID data and proposes FedDCG, a dynamic-consensus-graph-based method that integrates consensus-based weighting and sample-size-aware aggregation.

AIに質問

Bookmark

View Full Paper