Research map / Learning theory and algorithms / Kernels and regression theory
Optimality Implies Kernel Sum Classifiers are Statistically Efficient
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 KKT optimality to restrict analysis to optimal classifiers | Kernel sum SVM · Sum of m kernels | dual quadratic form: 3 m^{-\log_2(3)} B^2 | Helped |
| Use KKT optimality to restrict analysis to optimal classifiers | Kernel subset SVM · Kernel subsets | empirical Rademacher complexity bound: 2^{m-1} -> O\left(\frac{BRm^{0.208} \sqrt{\ln(m)}}{\sqrt n}\right) | Helped |
| Replace one kernel with a sum of m kernels | Optimal kernel SVM · Kernel sums | empirical Rademacher complexity: \frac{BR}{\sqrt n} -> O\left(\frac{BRm^{0.208}}{\sqrt n}\right) | Hurt |
| Allow selection of any subset of kernels rather than fixing one subset | Optimal kernel subset SVM · Kernel subsets | empirical Rademacher complexity: \sqrt{\ln(m)} | Hurt |
| Use an ℓ2-slack SVM with C=\frac12 | Two-kernel SVM · Sum of two kernels, non-separable data | dual quadratic form upper bound: \frac23 \max\{\valpha_1^\intercal\mKt_1\valpha_1, \valpha_2^\intercal\mKt_2\valpha_2\} | Helped |
Closest papers by meaning
- The optimality of kernel classifiers in Sobolev space (ICLR 2024 · similarity 0.88)
- Infinite Kernel Learning: Generalization Bounds and Algorithms (AAAI 2017 · similarity 0.88)
- On Contraction of Sequential and Offset Rademacher Complexities (ICML 2026 · similarity 0.87)
- Learning Kernels with Random Features (NeurIPS 2016 · similarity 0.87)
- Multi-Class Learning: From Theory to Algorithm (NeurIPS 2018 · similarity 0.86)
- Learning SMaLL Predictors (NeurIPS 2018 · 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.