Research map / Adversarial robustness and security / Certified and empirical robustness
Efficiently Learning Adversarially Robust Halfspaces with Noise
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use an efficient separation oracle for the perturbation set to solve robust empirical risk minimization | Halfspaces in the realizable setting · Realizable robust halfspace learning | runtime: poly(m,d,b) | Helped |
| Require an efficient approximate separation oracle to compute robust loss | Halfspace robust loss evaluation · Robust loss computation for convex perturbation sets | oracle requirement: poly(d,b,log(1/γ), log(R)) | Mixed |
| Relate ℓ_p-robust loss to γ-margin loss and optimize a convex surrogate | Halfspaces under random classification noise · Random classification noise, ℓ_p perturbations | robust misclassification error: η + ε | Helped |
| Use a hinge-like surrogate with noise-adjusted slopes | Halfspaces under random classification noise · Random classification noise, ℓ_p perturbations | robust misclassification error: η + ε | Helped |
Closest papers by meaning
- Provable Robustness of Adversarial Training for Learning Halfspaces with Noise (ICML 2021 · similarity 0.96)
- The Complexity of Adversarially Robust Proper Learning of Halfspaces with Agnostic Noise (NeurIPS 2020 · similarity 0.94)
- Learning General Halfspaces with Adversarial Label Noise via Online Gradient Descent (ICML 2022 · similarity 0.91)
- An Efficient Tester-Learner for Halfspaces (ICLR 2024 · similarity 0.87)
- Robust learning of halfspaces under log-concave marginals (NeurIPS 2025 · similarity 0.86)
- Learning Noisy Halfspaces with a Margin: Massart is No Harder than Random (NeurIPS 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.