Research map / Learning theory and algorithms / Prediction and decision losses
Energy-Efficient Scheduling with Predictions
Eric Balkanski, Noemie Perivier, Clifford Stein, Hao-Ting Wei
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 TPE-S with predictions instead of 2-competitive online algorithm at small error parameter | TPE-S · periodic, power-law and College Message datasets | competitive ratio (energy plus flow time) | Improved |
| increase error parameter sigma to large values | TPE-S · periodic, power-law and College Message datasets | competitive ratio | Hurt |
| increase shift tolerance eta_shift (initial phase) (ablation) | TPE-S · periodic dataset, sigma=0.4 | competitive ratio | Improved |
| increase shift tolerance eta_shift further (later phase) (ablation) | TPE-S · periodic dataset, sigma=0.4 | competitive ratio | Hurt |
| increase lambda (weight of the prediction-ignoring first phase) (ablation) | TPE-S · periodic dataset, sigma=0.4 | competitive ratio under small error | Hurt |
| increase lambda when predictions are incorrect (ablation) | TPE-S · periodic dataset | competitive ratio under large error | Improved |
Closest papers by meaning
- Learning Augmented Energy Minimization via Speed Scaling (NeurIPS 2020 · similarity 0.89)
- Competitive Fair Scheduling with Predictions (ICLR 2025 · similarity 0.87)
- What to Predict for Efficient Scheduling on Parallel Machines (NeurIPS 2026 · similarity 0.86)
- Non-clairvoyant Scheduling with Partial Predictions (ICML 2024 · similarity 0.85)
- Beyond Prediction: Tail-Aware Scheduling for LLM Inference (ICML 2026 · similarity 0.82)
- Learning-Augmented Priority Queues (NeurIPS 2024 · similarity 0.82)
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.
