Controlling Federated Learning for Covertness
Not on the map. Why?
Abstract (arXiv)
A learner aims to minimize a function f by repeatedly querying a distributed oracle that provides noisy gradient evaluations. At the same time, the learner seeks to hide argmin f from a malicious eavesdropper that observes the learner's queries. This paper considers the problem of textitcovert or textitlearner-private optimization, where the learner has to dynamically choose between learning and obfuscation by exploiting the stochasticity. The problem of controlling the stochastic gradient algorithm for covert optimization is modeled as a Markov decision process, and we show that the dynamic programming operator has a supermodular structure implying that the optimal policy has a monotone threshold structure. A computationally efficient policy gradient algorithm is proposed to search for the optimal querying policy without knowledge of the transition probabilities. As a practical application, our methods are demonstrated on a hate speech classification task in a federated setting where an eavesdropper can use the optimal weights to generate toxic content, which is more easily misclassified. Numerical results show that when the learner uses the optimal policy, an eavesdropper can only achieve a validation accuracy of 52% with no information and 69% when it has a public dataset with 10% positive samples compared to 83% when the learner employs a greedy policy.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 40 of its 70 references that are in the map. First posted on arXiv 17 Aug 2023.
Builds on
40 of 70 references.
References · 40
- 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
- 2014Adam: A Method for Stochastic Optimizationcited by 37,240
- 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
- 2019Advances and Open Problems in Federated Learningcited by 3,046
- 2013Stochastic First- and Zeroth-order Methods for Nonconvex Stochastic Programmingcited by 724
- 2018Federated Learning for Mobile Keyboard Predictioncited by 572
- 2019Making AI Forget You: Data Deletion in Machine Learningcited by 465
- 2022ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech Detectioncited by 422
- 2021Remember What You Want to Forget: Algorithms for Machine Unlearningcited by 371
- 2017The Marginal Value of Adaptive Gradient Methods in Machine Learningcited by 250
- 2020DeltaGrad: Rapid retraining of machine learning modelscited by 149
- 2018Transferring GANs: generating images from limited datacited by 134
28 more
- 2022ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally!cited by 134
- 2020A Principled Approach to Data Valuation for Federated Learningcited by 127
- 2020Sample-Efficient Optimization in the Latent Space of Deep Generative Models via Weighted Retrainingcited by 123
- 2021Entailment as Few-Shot Learnercited by 123
- 2023Stochastic Controlled Averaging for Federated Learning with Communication Compressioncited by 101
- 2018Tunability: Importance of Hyperparameters of Machine Learning Algorithmscited by 99
- 2018LAG: Lazily Aggregated Gradient for Communication-Efficient Distributed Learningcited by 96
- 2020On the Convergence of SGD with Biased Gradientscited by 83
- 2021Private Stochastic Convex Optimization: Optimal Rates in ℓ1 Geometrycited by 83
- 2019Finite-Sample Analysis of Nonlinear Stochastic Approximation with Applications in Reinforcement Learningcited by 67
- 2022FedVARP: Tackling the Variance Due to Partial Client Participation in Federated Learningcited by 64
- 2018On Markov Chain Gradient Descentcited by 63
- 2017Adversarial Ranking for Language Generationcited by 50
- 2021A Theory of Label Propagation for Subpopulation Shiftcited by 48
- 2022Annotation Error Detection: Analyzing the Past and Present for a More Coherent Futurecited by 45
- 2019Training language GANs from Scratchcited by 31
- 2017Learning Algorithms for Active Learningcited by 30
- 2019How You Act Tells a Lot: Privacy-Leakage Attack on Deep Reinforcement Learningcited by 29
- 2022Data Leakage in Federated Averagingcited by 26
- 2023Federated Learning under Heterogeneous and Correlated Client Availabilitycited by 18
- 2020Convergence Rates of Accelerated Markov Gradient Descent with Applications in Reinforcement Learningcited by 15
- 2021Deceptive Reinforcement Learning for Privacy-Preserving Planningcited by 14
- 2021Hate Speech Classifiers Learn Human-Like Social Stereotypescited by 12
- 2018Private Sequential Learningcited by 5
- 2019Query Complexity of Bayesian Private Learningcited by 4
- 2021Learner-Private Convex Optimizationcited by 2
- 2017Computing monotone policies for Markov decision processes: a nearly-isotonic penalty approachcited by 1
- 2022Quantifying Human Bias and Knowledge to guide ML models during Trainingcited by 1
Built on by
Citing papers per year: 2024: 2.
2 papers.