Research map / Bandits and online learning / Online convex optimization and games
Improved Kernel Alignment Regret Bound for Online Kernel Learning
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 POMDR with its OMD update and adaptive budget management instead of B(AO)₂KS | Online kernel learning · magic04, σ=0.5, ζ=2/3 | AMR (%): 22.04 ± 0.43 -> 16.17 ± 0.21 | Helped |
| Use POMDR with its OMD update and adaptive budget management instead of FOGD | Online kernel learning · w8a, σ=4, ζ=0.5 | AMR (%): 2.25 ± 0.04 -> 2.12 ± 0.06 | Helped |
| Use POMDR with its OMD update and adaptive budget management instead of SkeGD | Online kernel learning · mushrooms, σ=2, ζ=2/3 | AMR (%): 0.26 ± 0.02 -> 0.21 ± 0.03 | Helped |
| Increase Gaussian kernel width from σ=0.5 to σ=4 | POMDR · magic04 | AMR (%): 16.17 -> 23.26 | Hurt |
| Increase Gaussian kernel width from σ=4 to σ=64 | POMDR · w8a | AMR (%): 2.12 -> 3.12 | Hurt |
| Increase Gaussian kernel width from σ=0.5 to σ=2 | POMDR · mushrooms | AMR (%): 0.83 -> 0.21 | Helped |
| Use POMDR with OMD and adaptive learning rates instead of FOGD, NOGD, and SkeGD | Online kernel learning · magic04, w8a, and mushrooms | AMR (%) | Helped |
| Use partial example removal and retain an informed starting hypothesis rather than restart when the budget fills | POMDR versus B(AO)₂KS · magic04, w8a, and mushrooms | AMR (%) | Helped |
Closest papers by meaning
- On the Target-kernel Alignment: a Unified Analysis with Kernel Complexity (NeurIPS 2024 · similarity 0.89)
- Sharper Guarantees for Misspecified Kernelized Bandit Optimization (NeurIPS 2026 · similarity 0.86)
- From Average Sensitivity to Small-Loss Regret Bounds under Random-Order Model (NeurIPS 2026 · similarity 0.85)
- Dynamic Regret Reduces to Kernelized Static Regret (NeurIPS 2025 · similarity 0.85)
- Ahpatron: A New Budgeted Online Kernel Learning Machine with Tighter Mistake Bound (AAAI 2024 · similarity 0.85)
- Generalization in Kernel Regression Under Realistic Assumptions (ICML 2024 · similarity 0.84)
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.