Research map / Learning theory and algorithms / Kernels and regression theory
Synergy of Monotonic Rules
Vladimir Vapnik, Rauf Izmailov
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 |
|---|---|---|---|
| replace voting with synergy of 3 SVM kernels | 3-kernel SVM ensemble · Adult | test error (%): 20.07% -> 19.08% | Improved |
| replace voting with synergy of 3 SVM kernels | 3-kernel SVM ensemble · Covertype | test error (%): 27.83% -> 28.96% | Hurt |
| replace voting with synergy of 3 SVM kernels | 3-kernel SVM ensemble · Australian | test error (%): 12.02% -> 12.54% | Hurt |
| replace voting with synergy of 3 SVM kernels | 3-kernel SVM ensemble · MONK's-2 | test error (%): 19.31% -> 16.23% | Improved |
| replace voting with synergy of 3 SVM kernels | 3-kernel SVM ensemble · MONK's-1 | test error (%): 22.80% -> 20.16% | Improved |
| replace voting with synergy of 3 SVM kernels | 3-kernel SVM ensemble · Tic-tac-toe | test error (%): 1.95% -> 1.75% | Improved |
| replace voting with synergy of 3 SVM kernels | 3-kernel SVM ensemble · Bank | test error (%): 12.79% -> 11.73% | Improved |
| synergy of RBF and INK-Spline trained on 300 samples vs single SVM on 3000 | Synergy on Adult · Adult | test error (%): 18.49% -> 17.92% | Improved |
| synergy of three RBF SVMs trained on disjoint 300-sample subsets vs voting | RBF SVM ensemble · Adult | test error (%): 19.44% -> 18.52% | Improved |
| synergy of 3 disjoint-subset SVMs vs one SVM on 900 samples | RBF SVM · Adult | test error (%): 20.01% -> 18.52% | Improved |
| replace argmax of OVR scores with synergy conditional probability | one-vs-rest SVM · Vehicle | test error (%): 17.45% -> 14.15% | Improved |
| replace argmax of OVR scores with synergy conditional probability | one-vs-rest SVM · Waveform | test error (%): 20.10% -> 18.31% | Improved |
Closest papers by meaning
- Fast and Flexible Monotonic Functions with Ensembles of Lattices (NeurIPS 2016 · similarity 0.83)
- A universally consistent learning rule with a universally monotone error (JMLR 2022 · similarity 0.82)
- Monotonic Calibrated Interpolated Look-Up Tables (JMLR 2016 · similarity 0.82)
- Smooth Min-Max Monotonic Networks (ICML 2024 · similarity 0.80)
- Robustly Learning Monotone Single-Index Models (NeurIPS 2025 · similarity 0.80)
- On the Consistency of Kernel Methods with Dependent Observations (ICML 2024 · similarity 0.80)
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.
