Adversarial robustness for latent models: revisiting the robust-standard accuracies tradeoff

DOI10.1287/OPRE.2022.0162MaRDI QIDQ6580508FDOQ6580508

Authors: A. Javanmard 

Publication date: 29 July 2024

Published in: Operations Research (Search for Journal in Brave)

Recommendations

• Precise statistical analysis of classification accuracies for adversarial training
• Robustness should not be at odds with accuracy
• LADDER: latent boundary-guided adversarial training
• Analysis of classifiers' robustness to adversarial perturbations
• The curse of overparametrization in adversarial training: precise analysis of robust generalization for random features regression

zbMATH Keywords

classificationlow-dimensional structuresadversarial trainingrobust machine learningmachine learning and data science

Mathematics Subject Classification ID

Robustness in mathematical programming (90C17)

Cited In (2)

• Robustness should not be at odds with accuracy
• Improving the accuracy-robustness trade-off of classifiers via adaptive smoothing

This page was built for publication: Adversarial robustness for latent models: revisiting the robust-standard accuracies tradeoff

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6580508)