Protecting Federated Learning from Extreme Model Poisoning Attacks via Multidimensional Time Series Anomaly Detection
Not on the map. Why?
Abstract (arXiv)
Current defense mechanisms against model poisoning attacks in federated learning (FL) systems have proven effective up to a certain threshold of malicious clients. In this work, we introduce FLANDERS, a novel pre-aggregation filter for FL resilient to large-scale model poisoning attacks, i.e., when malicious clients far exceed legitimate participants. FLANDERS treats the sequence of local models sent by clients in each FL round as a matrix-valued time series. Then, it identifies malicious client updates as outliers in this time series by comparing actual observations with estimates generated by a matrix autoregressive forecasting model maintained by the server. Experiments conducted in several non-iid FL setups show that FLANDERS significantly improves robustness across a wide spectrum of attacks when paired with standard and robust existing aggregation methods.
Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 27 of its 49 references that are in the map. First posted on arXiv 29 Mar 2023.
Builds on
27 of 49 references.
References · 27
- 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,592
- 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,283
- 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,827
- 2016Federated Learning: Strategies for Improving Communication Efficiencycited by 1,461
- 2019Measuring the Effects of Non-Identical Data Distribution for Federated Visual Classificationcited by 836
- 2018How To Backdoor Federated Learningcited by 745
- 2018Byzantine-Robust Distributed Learning: Towards Optimal Statistical Ratescited by 698
- 2019Local Model Poisoning Attacks to Byzantine-Robust Federated Learningcited by 479
- 2020Flower: A Friendly Federated Learning Research Frameworkcited by 449
- 2018Analyzing Federated Learning through an Adversarial Lenscited by 346
- 2020FLTrust: Byzantine-robust Federated Learning via Trust Bootstrappingcited by 330
- 2018The Hidden Vulnerability of Distributed Learning in Byzantiumcited by 324
15 more
- 2019A Little Is Enough: Circumventing Defenses For Distributed Learningcited by 285
- 2018Manipulating Machine Learning: Poisoning Attacks and Countermeasures for Regression Learningcited by 219
- 2020Multivariate Time-series Anomaly Detection via Graph Attention Networkcited by 165
- 2022FLDetector: Defending Federated Learning Against Model Poisoning Attacks via Detecting Malicious Clientscited by 119
- 2021Back to the Drawing Board: A Critical Evaluation of Poisoning Attacks on Production Federated Learningcited by 116
- 2020Neuron Shapley: Discovering the Responsible Neuronscited by 98
- 2018Generalized Byzantine-tolerant SGDcited by 94
- 2018Is feature selection secure against training data poisoning?cited by 88
- 2017Being Robust (in High Dimensions) Can Be Practicalcited by 83
- 2022Survey on Federated Learning Threats: concepts, taxonomy on attacks and defences, experimental study and challengescited by 81
- 2016Data Poisoning Attacks on Factorization-Based Collaborative Filteringcited by 78
- 2021Challenges and Approaches for Mitigating Byzantine Attacks in Federated Learningcited by 46
- 2018Autoregressive Models for Matrix-Valued Time Seriescited by 18
- 2022Blades: A Unified Benchmark Suite for Byzantine Attacks and Defenses in Federated Learningcited by 7
- 2021Turning Federated Learning Systems Into Covert Channelscited by 6
Built on by
Citing papers per year: 2023: 1; 2024: 0; 2025: 1; 2026: 1.
3 papers.