Research map / Learning theory and algorithms / Sample complexity
On the Hardness of Robust Classification
Pascale Gourdeau, Varun Kanade, Marta Kwiatkowska, James Worrell
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Require robustness to perturbations of ω(log n) bits | MON-CONJ under the uniform distribution · Uniform distribution | robust learnability | Hurt |
| Restrict the perturbation budget to O(log n) | MON-CONJ under α-log-Lipschitz distributions · α-log-Lipschitz distributions | robust learnability | Improved |
| Require distribution-free robustness against one-bit perturbations | Non-trivial concept classes · Distribution-free on {0,1}^n | robust learnability | Hurt |
| Encode each input bit in repeated blocks and use a majority decoder | Constructed robust-learning class · Constructed concept-distribution pairs | sample complexity: O\left(\frac{1}{\epsilon}\left(\log\abs{\mathcal{C}_n}+\log\frac{1}{\delta}\right)\right) | Improved |
Closest papers by meaning
- Adversarial examples from computational constraints (ICML 2019 · similarity 0.89)
- Sample Complexity of Robust Linear Classification on Separated Data (ICML 2021 · similarity 0.88)
- The Complexity of Adversarially Robust Proper Learning of Halfspaces with Agnostic Noise (NeurIPS 2020 · similarity 0.88)
- Adversarially Robust Generalization Requires More Data (NeurIPS 2018 · similarity 0.87)
- Gradient Flow Provably Learns Robust Classifiers for Orthonormal GMMs (ICML 2025 · similarity 0.87)
- Probably Approximately Global Robustness Certification (ICML 2025 · similarity 0.87)
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.
