Research map / Adversarial robustness and security / Certified and empirical robustness
Towards Consistency in Adversarial Classification
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 a convex margin surrogate at positive adversarial radius | adversarial classification · adversarial classification | adversarial calibration | Hurt |
| shift an odd margin loss by a positive offset | odd margin losses · adversarial classification | adversarial calibration | Helped |
| use a standard-calibrated margin loss on an ε-realisable distribution | adversarial classification · ε-realisable distributions | adversarial consistency | Helped |
| use a 0/1-like margin loss | adversarial classification · distributions on spaces with the midpoint property | optimal adversarial risk | No clear effect |
Closest papers by meaning
- The Adversarial Consistency of Surrogate Risks for Binary Classification (NeurIPS 2023 · similarity 0.93)
- Calibration and Consistency of Adversarial Surrogate Losses (NeurIPS 2021 · similarity 0.91)
- Adversarial Multiclass Classification: A Risk Minimization Perspective (NeurIPS 2016 · similarity 0.88)
- Theoretically Principled Trade-off between Robustness and Accuracy (ICML 2019 · similarity 0.88)
- Expressive Losses for Verified Robustness via Convex Combinations (ICLR 2024 · similarity 0.88)
- Consistent Adversarially Robust Linear Classification: Non-Parametric Setting (ICML 2024 · similarity 0.88)
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.