Research map / Learning theory and algorithms / Sample complexity
On the Power of Localized Perceptron for Label-Optimal Learning of Halfspaces with Adversarial 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 the proposed localized online active learner instead of the prior passive learner | Halfspace learning with isotropic log-concave marginals and adversarial noise · Isotropic log-concave halfspaces, adversarial noise | label complexity: \tilde{O}(d/ \epsilon^4) -> \tilde{O}(d \cdot \operatorname{polylog}(\frac{1}{\epsilon})) | Helped |
| Use the proposed localized online active learner instead of the prior active learner | Halfspace learning under adversarial noise · Uniform halfspaces, adversarial noise | noise tolerance: \nu = \tilde{\Omega}(\epsilon/ \log d) -> \nu = \Omega(\epsilon) | Helped |
| Use the proposed learner on isotropic log-concave marginals rather than the prior active learner restricted to uniform marginals | Online active halfspace learning · Isotropic log-concave halfspaces, adversarial noise | supported marginal distributions | Helped |
| Use prior knowledge of s-sparsity with the proposed learner | Localized Perceptron · Isotropic log-concave s-sparse halfspaces, adversarial noise | label complexity: \tilde{O}(d \cdot \operatorname{polylog}(\frac{1}{\epsilon})) -> \tilde{O}(s \cdot \operatorname{polylog}(d, \frac{1}{\e | Helped |
| Translate adversarial-noise learning to the agnostic model | Localized Perceptron · Agnostic halfspace learning | error rate: c \cdot \mathrm{OPT} + \epsilon | Helped |
Closest papers by meaning
- An Efficient Tester-Learner for Halfspaces (ICLR 2024 · similarity 0.90)
- Active Learning of General Halfspaces: Label Queries vs Membership Queries (NeurIPS 2024 · similarity 0.89)
- Learning Noisy Halfspaces with a Margin: Massart is No Harder than Random (NeurIPS 2024 · similarity 0.88)
- A Near-optimal Algorithm for Learning Margin Halfspaces with Massart Noise (NeurIPS 2024 · similarity 0.87)
- Active Classification with Few Queries under Misspecification (NeurIPS 2024 · similarity 0.86)
- Online Linear Classification with Massart Noise (ICML 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.