Attack-Resistant Federated Learning with Residual-based Reweighting

Shuhao Fu, Chulin Xie, Bo Li, Qifeng Chen
12/24/2019
cs.LGstat.ML
federated

Official code: fushuhao6/Attack-Resistant-Federated-Learning (32 stars, last commit 2020-04-08).

Abstract

Federated learning has a variety of applications in multiple domains by utilizing private training data stored on different devices. However, the aggregation process in federated learning is highly vulnerable to adversarial attacks so that the global model may behave abnormally under attacks. To tackle this challenge, we present a novel aggregation algorithm with residual-based reweighting to defend federated learning. Our aggregation algorithm combines repeated median regression with the reweighting scheme in iteratively reweighted least squares. Our experiments show that our aggregation algorithm outperforms other alternative algorithms in the presence of label-flipping and backdoor attacks. We also provide theoretical analysis for our aggregation algorithm.

View on arXivView PDF

Code Implementations(1)

fushuhao6/Attack-Resistant-Federated-LearningOfficial100%
324Shell, PythonJan 21, 20206 years ago

Similar Papers

Robust Aggregation for Federated Learning

86% match1 repo532019

DFedReweighting: A Unified Framework for Objective-Oriented Reweighting in Decentralized Federated Learning

86% match2025

FedSECA: Sign Election and Coordinate-wise Aggregation of Gradients for Byzantine Tolerant Federated Learning

86% match1 repo22024

A Whole-Process Certifiably Robust Aggregation Method Against Backdoor Attacks in Federated Learning

85% match2024

FedRDF: A Robust and Dynamic Aggregation Function against Poisoning Attacks in Federated Learning

85% match2024

A Secure Aggregation for Federated Learning on Long-Tailed Data

85% match2023

Cite this paper

@article{fu2019attackresistant,
  title  = {Attack-Resistant Federated Learning with Residual-based Reweighting},
  author = {Shuhao Fu and Chulin Xie and Bo Li and Qifeng Chen},
  year   = {2019},
  eprint = {1912.11464},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1912.11464}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1912.11464.svg)](https://codeofpaper.com/paper/1912.11464)

Discussion