Research map / Adversarial robustness and security / Certified and empirical robustness
Adversarially Robust PAC Learnability of Real-Valued Functions
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 median aggregation of weak learners (median boosting) instead of averaging strong base learners | robust l_p regression ensemble · sample complexity | sample complexity exponent | Helped |
| Sparsify the boosted ensemble to k=Õ(fat*(H,cε)) functions | median boosting ensemble · robust l_p regression | compression size | Helped |
| Convex function classes: the averaged ensemble output is a convex combination of class members, so the learner is proper | robust l_p regression ensemble · convex classes | proper learnability | Helped |
| Agnostic learner for robust (eta,beta)-regression with a fixed cutoff applied to robust regression (changing cutoffs) | robust regression l_p · robust regression | error guarantee | No clear effect |
Closest papers by meaning
- Reducing Adversarially Robust Learning to Non-Robust PAC Learning (NeurIPS 2020 · similarity 0.90)
- On Proper Learnability between Average- and Worst-case Robustness (NeurIPS 2023 · similarity 0.90)
- A Characterization of Semi-Supervised Adversarially Robust PAC Learnability (NeurIPS 2022 · similarity 0.90)
- On Robustness to Adversarial Examples and Polynomial Optimization (NeurIPS 2019 · similarity 0.88)
- The Complexity of Adversarially Robust Proper Learning of Halfspaces with Agnostic Noise (NeurIPS 2020 · similarity 0.88)
- Algorithms and SQ Lower Bounds for Robustly Learning Real-valued Multi-Index Models (NeurIPS 2025 · 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.