Research map / Bandits and online learning / Online convex optimization and games
Dynamic Local Regret for Non-convex Online Forecasting
Sergul Aydore, Tianhao Zhu, Dean P. Foster
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 |
|---|---|---|---|
| Use dynamic exponentially time-smoothed SGD instead of online SGD | LSTM · GEFCom2014 electrical loads, test set | QL_grand: 14,612 -> 14,585 | Improved |
| Use dynamic exponentially time-smoothed SGD instead of static time-smoothed SGD | LSTM · GEFCom2014 electrical loads, test set | QL_grand: 14,595 -> 14,585 | Improved |
| Use dynamic exponentially time-smoothed SGD rather than online SGD | LSTM · GEFCom2014 electrical loads, varying learning rates | robustness to learning-rate selection | Improved |
| Use dynamic exponentially time-smoothed SGD rather than static time-smoothed SGD | LSTM · GEFCom2014 electrical loads, varying window sizes and learning rates | stability | Improved |
| Use dynamic exponentially time-smoothed SGD rather than offline SGD retrained on all past data | LSTM · GEFCom2014 electrical loads, varying window sizes | computation time | Improved |
| Evaluate historical losses at their corresponding historical parameters rather than at the current parameter | Dynamic Local Regret · Toy example, T=3 | local regret ranking of oracle versus stale policy: Static Local Regret: $40/9$ vs. $4/9$; Dynamic Local Regret: 0 vs. $10/9$ | Improved |
| Increase the STS-SGD window size | STS-SGD · GEFCom2014 electrical loads, varying learning rates | stability | Hurt |
| Increase the DTS-SGD window size | DTS-SGD · GEFCom2014 electrical loads, varying learning rates | stability | No clear effect |
Closest papers by meaning
- Online Linear Regression in Dynamic Environments via Discounting (ICML 2024 · similarity 0.88)
- On the Dynamic Regret of Following the Regularized Leader: Optimism with History Pruning (ICML 2025 · similarity 0.88)
- Non-stationary Online Learning for Curved Losses: Improved Dynamic Regret via Mixability (ICML 2025 · similarity 0.88)
- Dynamic Regret in Online Convex Optimization with Indicator Switching Costs (NeurIPS 2026 · similarity 0.87)
- Dynamic Regret Reduces to Kernelized Static Regret (NeurIPS 2025 · similarity 0.87)
- An Equivalence Between Static and Dynamic Regret Minimization (NeurIPS 2024 · similarity 0.87)
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.
