Machine Learning with Adversaries: Byzantine Tolerant Gradient Descent

By Peva Blanchard, El Mahdi El Mhamdi, Rachid Guerraoui, Julien Stainer. 2017

2,816 citations recorded by the source corpus.

Subfields: Safety, Robustness & Interpretability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Safety, Robustness & Interpretability

Citations in the corpus

Benchmarking Neural Network Robustness to Common Corruptions and Perturbations (2019, background): “, 2018), and model or data poisoning (Blanchard et al., 2017; Steinhardt et al., 2017; Hendrycks et al., 2018).”
Advances and Open Problems in Federated Learning (2019, background): “ntargeted model update poisoning attacks, especially Byzantine attacks, replaces the averaging step on the server with a robust estimate of the mean, such as median-based aggregators [101, 426], Krum [69], and trimmed mean [426]. Past work has shown that various robust aggregators are provably effec”
How To Backdoor Federated Learning (2018, methodology): “Recent work [5, 13, 14, 71] proposed alternative aggregation mechanisms to ensure convergence (but not integrity) in the presence of Byzantine participants.”
Local Model Poisoning Attacks to Byzantine-Robust Federated Learning (2019, methodology): “In particular, an attacker can manipulate the global model parameters arbitrarily for this mean aggregation rule when compromising only one worker device [9,66].”
FedML: A Research Library and Benchmark for Federated Machine Learning (2020, methodology): “K RUM [119].”
IBM Federated Learning: an Enterprise Framework White Paper V0.1 (2020, background): “Algorithms to improve robustness such as Krum [7], a Byzantine tolerant gradient descent coordinate-wise median fusion [14], and Zeno, a distributed stochastic gradient descent with suspicion-based fault-tolerance [32] are also available.”
Attack of the Tails: Yes, You Really Can Backdoor Federated Learning (2020, methodology): “Although MULTI-KRUM selects model updates from clients with Southwest Airlines examples, it is due to it rejecting few model updates per FL round i.e., when multiple honest clients with Southwest examples appear in the same FL round, MULTI-KRUM will not reject all of their model updates.”
Backdoor Attacks on Federated Meta-Learning (2020, methodology): “The same assumption was made in [12, 13, 19, 16, 17, 20, 21], where outliers were detected and removed according to slightly different measures taken from the distribution of benign values (benign users were assumed to be the majority).”
Salvaging Federated Learning by Local Adaptation (2020, methodology): “To limit the influence of individual participants on the federated model, several robust, “Byzantine-tolerant” aggregation schemes have been proposed (Blanchard et al., 2017; El Mhamdi et al., 2018; Damaskinos et al., 2019; Rajput et al., 2019; Chen et al., 2017).”
Turbo-Aggregate: Breaking the Quadratic Aggregation Barrier in Secure Federated Learning (2020, background): “ifying their local datasets. The setup when adversaries can modify their local datasets to inﬂuence the global model has been studied in the literature in the context of non-secure federated learning [49], [50]. These frameworks often rely on comparing the distances between the local updates of the ”
Robust Aggregation for Federated Learning (2019, methodology): “Byzantine robustness, resilience to arbitrary behavior of some devices [51], was studied in distributed optimization with gradient aggregation [5, 16, 20, 21, 23, 88].”
Can You Really Backdoor Federated Learning? (2019, background): “Under untargeted attacks (6; 15; 12), the goal of the adversary is to corrupt the model in such a way that it does not achieve a near-optimal performance on the main task at hand (e.”
DETOX: A Redundancy-based Framework for Faster and More Robust Gradient Aggregation (2019, methodology): “Unfortunately, the presence of a single Byzantine compute node can result in arbitrarily bad global models when aggregating gradients via their average [1].”
Data Encoding for Byzantine-Resilient Distributed Optimization (2019, background): “Distributed computing with Byzantine adversaries is a richly investigated topic since [LSP82], and has received recent attention in the context of large-scale distributed optimization and learning [CSX17,CWCP18,YCRB18,BMGS17].”
Zeno++: Robust Fully Asynchronous SGD (2019, background): “In particular, tolerance to Byzantine failures (Blanchard et al., 2017; Chen et al., 2017; Yin et al., 2018; Feng et al., 2014; Su & Vaidya, 2016; 2018; Xie et al., 2019b; Alistarh et al., 2018; Cao & Lai, 2018; Xie et al., 2019c) has become an important topic in the distributed machine learning…”
Analyzing Federated Learning through an Adversarial Lens (2018, methodology): “In Section 4, we study the effect of our attacks on the Byzantine-resilient aggregation mechanisms ‘Krum’ (Blanchard et al., 2017) and coordinate-wise median (Yin et al., 2018).”
Distributed Learning over Unreliable Networks (2018, background): “Also, in Blanchard et al. (2017); Yin et al. (2018); Alistarh et al. (2018) They study the Byzantine-robust distributed learning when Byzantine worker included in the network.”
Mitigating Sybils in Federated Learning Poisoning (2018, methodology): “However, state of the art magnitude-based detection methods exist [9] to prevent these attacks, and we also do not consider them in our solution.”