mlsanityData to 7 Oct 2026

On the Role of Randomization in Adversarially Robust Classification

arXiv 2302.07221 ↗First posted 14 Feb 2023cs.LGcited by 4

Not on the map. Why?

Abstract (arXiv)

Deep neural networks are known to be vulnerable to small adversarial perturbations in test data. To defend against adversarial attacks, probabilistic classifiers have been proposed as an alternative to deterministic ones. However, literature has conflicting findings on the effectiveness of probabilistic classifiers in comparison to deterministic ones. In this paper, we clarify the role of randomization in building adversarially robust classifiers. Given a base hypothesis set of deterministic classifiers, we show the conditions under which a randomized ensemble outperforms the hypothesis set in adversarial risk, extending previous results. Additionally, we show that for any probabilistic binary classifier (including randomized ensembles), there exists a deterministic classifier that outperforms it. Finally, we give an explicit description of the deterministic hypothesis set that contains such a deterministic classifier for many types of commonly used probabilistic classifiers, i.e. randomized ensembles and parametric/input noise injection.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 39 of its 53 references that are in the map. First posted on arXiv 14 Feb 2023.

Builds on

39 of 53 references.

References · 39

  1. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  2. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  3. 2013Intriguing properties of neural networkscited by 5,114
  4. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,442
  5. 2019Certified Adversarial Robustness via Randomized Smoothingcited by 1,324
  6. 2016Adversarial Machine Learning at Scalecited by 934
  7. 2017Evasion Attacks against Machine Learning at Test Timecited by 686
  8. 2017Adversarial Examples Are Not Easily Detected: Bypassing Ten Detection Methodscited by 542
  9. 2018Certified Robustness to Adversarial Examples with Differential Privacycited by 462
  10. 2019On Evaluating Adversarial Robustnesscited by 442
  11. 2017Mitigating Adversarial Effects Through Randomizationcited by 382
  12. 2019Provably Robust Deep Learning via Adversarially Trained Smoothed Classifierscited by 328
27 more
  1. 2019Improving Adversarial Robustness via Promoting Ensemble Diversitycited by 213
  2. 2018Certified Adversarial Robustness with Additive Noisecited by 178
  3. 2018Robustness via curvature regularization, and vice versacited by 174
  4. 2018Stochastic Activation Pruning for Robust Adversarial Defensecited by 171
  5. 2017Towards Robust Neural Networks via Random Self-ensemblecited by 158
  6. 2018Parametric Noise Injection: Trainable Randomness to Improve Deep Neural Network Robustness against Adversarial Attackcited by 123
  7. 2020DVERGE: Diversifying Vulnerabilities for Enhanced Robust Generation of Ensemblescited by 73
  8. 2019Improving Adversarial Robustness of Ensembles with Diversity Trainingcited by 69
  9. 2020Randomization matters. How to defend against strong adversarial attackscited by 37
  10. 2021TRS: Transferability Reduced Ensemble via Encouraging Gradient Diversity and Model Smoothnesscited by 37
  11. 2019Theoretical evidence for adversarial robustness through randomizationcited by 33
  12. 2018Adversarial Risk and Robustness: General Definitions and Implications for the Uniform Distributioncited by 32
  13. 2022The Many Faces of Adversarial Riskcited by 28
  14. 2021On the Existence of the Adversarial Bayes Classifier (Extended Version)cited by 25
  15. 2019On the Hardness of Robust Classificationcited by 23
  16. 2019Ensembles of Locally Independent Prediction Modelscited by 22
  17. 2021Mixed Nash Equilibria in the Adversarial Examples Gamecited by 22
  18. 2022The Multimarginal Optimal Transport Formulation of Adversarial Multiclass Classificationcited by 21
  19. 2018ResNets Ensemble via the Feynman-Kac Formalism to Improve Natural and Robust Accuraciescited by 18
  20. 2022Existence and Minimax Theorems for Adversarial Surrogate Risks in Binary Classificationcited by 15
  21. 2021Local Competition and Stochasticity for Adversarial Robustness in Deep Learningcited by 12
  22. 2021Stochastic Local Winner-Takes-All Networks Enable Profound Adversarial Robustnesscited by 11
  23. 2021On the robustness of randomized classifiers to adversarial examplescited by 9
  24. 2022Building Robust Ensembles via Margin Boostingcited by 8
  25. 2019Robust Attacks against Multiple Classifierscited by 7
  26. 2022Adversarial Vulnerability of Randomized Ensemblescited by 5
  27. 2023On the Robustness of Randomized Ensembles to Adversarial Perturbationscited by 3

Built on by

Cited by year

Citing papers per year: 2024: 3; 2025: 1.

4 papers.

  1. 2024A Notion of Uniqueness for the Adversarial Bayes Classifiercited by 3
  2. 2024Adversarial Consistency and the Uniqueness of the Adversarial Bayes Classifiercited by 2
  3. 2024Uniform Convergence of Adversarially Robust Classifierscited by 1
  4. 2025Unveiling the Role of Randomization in Multiclass Adversarial Classification: Insights from Graph Theorycited by 0