Research map / Learning theory and algorithms / Kernels and regression theory
Multiclass learning with margin: exponential rates with no bias-variance trade-off
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 square loss instead of logistic or exponential loss | RFF linear model · Synthetic hard-margin classification, δ = 0.1 and δ = 0.2 | 0-1 error convergence | Helped |
| Optimize a surrogate loss by gradient descent under a hard margin | RFF linear model · Synthetic hard-margin classification | 0-1 error convergence relative to surrogate-loss convergence | Helped |
| Increase the soft-margin exponent α | Regularized logistic-loss linear model · Synthetic soft-margin classification | 0-1 error convergence with dataset size: n^{-α/2} | Helped |
| Fit error-rate slopes across increasing α | Regularized logistic-loss linear model · Synthetic soft-margin classification | slope of error-rate slopes versus α: -0.35; -0.5 | Mixed |
Closest papers by meaning
- Active Learning of Classifiers with Label and Seed Queries (NeurIPS 2022 · similarity 0.87)
- Polynomial-Time Robust Multiclass Linear Classification under Gaussian Marginals (NeurIPS 2026 · similarity 0.86)
- Trading off Consistency and Dimensionality of Convex Surrogates for Multiclass Classification (NeurIPS 2024 · similarity 0.86)
- A Universal Growth Rate for Learning with Smooth Surrogate Losses (NeurIPS 2024 · similarity 0.86)
- The Implicit Bias of Gradient Descent on Separable Multiclass Data (NeurIPS 2024 · similarity 0.86)
- Balancing the Scales: A Theoretical and Algorithmic Framework for Learning from Imbalanced Data (ICML 2025 · similarity 0.85)
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.