Research map / Bandits and online learning / Online convex optimization and games
Towards Optimal Robustness in Learning-Augmented Paging
Peng Chen, Hailiang Zhao, Xueyan Tang, Yixuan Wang, Shuiguang Deng
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 |
|---|---|---|---|
| add relative prediction budget (RPB) gating to ONOPT-OM (ablation) | ONOPT-OM · SPEC CPU 2006 avg (PLECO predictor) | cost ratio: 1.253 -> 1.249 | Improved |
| add relative prediction budget (RPB) gating to ONOPT-OM (ablation) | ONOPT-OM · SPEC CPU 2006 avg (POPU predictor) | cost ratio: 1.237 -> 1.220 | Improved |
| increase budget reset value tau from 1 to 4 (ablation) | RPB-OM · SPEC CPU 2006 avg (POPU predictor) | cost ratio: 1.220 -> 1.213 | Improved |
| increase budget reset value tau from 1 to 4 (ablation) | RPB-OM · SPEC CPU 2006 avg (PLECO predictor) | cost ratio: 1.249 -> 1.254 | No clear effect |
| use RPB-OM instead of BlindOracle&LRU | learning-augmented paging · SPEC CPU 2006 avg (PLECO predictor) | cost ratio: 1.286 -> 1.249 | Improved |
| use RPB-OM instead of LMarker (conservative budget earning, not 1-consistent) | learning-augmented paging · SPEC CPU 2006 avg (POPU predictor) | cost ratio: 1.320 -> 1.220 | Improved |
| add predictions to OnlineMin (RPB-OM vs OM) | OnlineMin · SPEC CPU 2006 avg (PLECO predictor) | cost ratio: 1.396 -> 1.249 | Improved |
Closest papers by meaning
- Overcoming Brittleness in Pareto-Optimal Learning Augmented Algorithms (NeurIPS 2024 · similarity 0.86)
- Online Weighted Paging with Unknown Weights (NeurIPS 2024 · similarity 0.84)
- Robustifying Learning-Augmented Caching Efficiently without Compromising 1-Consistency (NeurIPS 2025 · similarity 0.83)
- Statistical Properties of Robust Satisficing (ICML 2024 · similarity 0.83)
- Near-Optimal Consistency-Robustness Trade-Offs for Learning-Augmented Online Knapsack Problems (ICML 2025 · similarity 0.83)
- Robust Learning-Augmented Dictionaries (ICML 2024 · similarity 0.83)
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.
