Research map / Bandits and online learning / Online convex optimization and games
Robust Learning for Smoothed Online Convex Optimization with Feedback Delay
Pengfei Li, Jianyi Yang, Adam Wierman, Shaolei Ren
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 |
|---|---|---|---|
| train the ML model in a robustification-aware manner (RCL vs RCL-O), lambda=1 (ablation) | RCL · EV battery SOCO (ElaadNL), no delay | average cost: 1.2432 -> 1.1144 | Improved |
| train the ML model in a robustification-aware manner (RCL vs RCL-O), lambda=1 (ablation) | RCL · EV battery SOCO (ElaadNL), no delay | competitive ratio: 2.4209 -> 1.2905 | Improved |
| increase lambda from 0.6 to 1 | RCL · EV battery SOCO (ElaadNL), no delay | average cost: 1.4704 -> 1.1144 | Improved |
| increase lambda from 0.6 to 1 | RCL · EV battery SOCO (ElaadNL), no delay | competitive ratio: 1.7672 -> 1.2905 | Improved |
| increase lambda from 1 to 5 | RCL · EV battery SOCO (ElaadNL), no delay | average cost: 1.1144 -> 1.0441 | Improved |
| increase lambda from 1 to 3 | RCL · EV battery SOCO (ElaadNL), no delay | competitive ratio: 1.2905 -> 1.4405 | Hurt |
| increase lambda from 3 to 5 | RCL · EV battery SOCO (ElaadNL), no delay | competitive ratio: 1.4405 -> 1.3014 | Improved |
| add robustification (RCL lambda=5 vs plain ML) | ML · EV battery SOCO (ElaadNL), no delay | average cost: 1.0668 -> 1.0441 | Improved |
| add robustification to ML (RCL lambda=1 vs ML) | ML · EV battery SOCO (ElaadNL), no delay | competitive ratio: 3.2566 -> 1.2905 | Improved |
| swap expert ROBD for HitMin (ablation) | RCL lambda=1 · EV battery SOCO (ElaadNL), no delay | average cost: 1.1144 -> 1.0515 | Improved |
| swap expert ROBD for HitMin (ablation) | RCL lambda=1 · EV battery SOCO (ElaadNL), no delay | competitive ratio: 1.2905 -> 1.6035 | Hurt |
| robustification-aware training under 1-step feedback delay (ablation) | RCL · EV battery SOCO (ElaadNL), 1-step feedback delay | average cost: 1.3690 -> 1.3594 | Improved |
Closest papers by meaning
- A Reduction from Delayed to Immediate Feedback for Online Convex Optimization with Improved Guarantees (NeurIPS 2026 · similarity 0.88)
- Gradient-Variation Online Adaptivity for Accelerated Optimization with Hölder Smoothness (NeurIPS 2025 · similarity 0.88)
- Unconstrained Robust Online Convex Optimization (ICML 2025 · similarity 0.87)
- Capacity-Constrained Online Convex Optimization with Delayed Feedback (NeurIPS 2026 · similarity 0.87)
- Exploiting Curvature in Online Convex Optimization with Delayed Feedback (ICML 2025 · similarity 0.86)
- Safe Online Convex Optimization with Heavy-Tailed Observation Noises (AAAI 2025 · 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.
