Research map / Time series / Foundation models for time series
Zero-shot Forecasting by Simulation Alone
Boris N. Oreshkin, Mayank Jauhari, Ravi Kiran Selvam, Malcolm Wolff, Wenhao Pan, Shankar Ramasubramanian, Kin G. Olivares, Tatiana Konstantinova, Andres Potapczynski, Mengfei Cao, Dmitry Efimov, Michael W. Mahoney and 1 more
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 SARIMA-2 multi-seasonality superposition (ablation) | PatchTST-SarSim0 (250K steps) · GiftEval | sCRPS: 0.647 -> 0.576 | Improved |
| add SARIMA-2 multi-seasonality superposition (ablation) | NBEATS-SarSim0 · GiftEval | sCRPS: 0.655 -> 0.602 | Improved |
| add SARIMA-2 multi-seasonality superposition (ablation) | NBEATS-SarSim0 · M-Series | MASE: 0.941 -> 0.869 | Improved |
| add heavy-tailed Noisers (ablation) | PatchTST-SarSim0 (250K steps) · GiftEval | sCRPS: 0.594 -> 0.576 | Improved |
| add heavy-tailed Noisers (ablation) | PatchTST-SarSim0 (250K steps) · M-Series | MASE: 0.861 -> 0.922 | Hurt |
| add heavy-tailed Noisers (ablation) | NBEATS-SarSim0 · GiftEval | sCRPS: 0.609 -> 0.602 | No clear effect |
| add heavy-tailed Noisers (ablation) | NBEATS-SarSim0 · M-Series | MASE: 0.860 -> 0.869 | No clear effect |
| train 500K instead of 250K steps (ablation) | PatchTST-SarSim0 · M-Series | MASE: 0.922 -> 0.877 | Improved |
| train 500K instead of 250K steps (ablation) | PatchTST-SarSim0 · GiftEval | sCRPS: 0.576 -> 0.573 | No clear effect |
| pretrain on SarSim0 instead of KernelSynth data | PatchTST · GiftEval | sCRPS: 0.702 -> 0.573 | Improved |
| pretrain on SarSim0 instead of ForecastPFN data | PatchTST · GiftEval | sCRPS: 1.060 -> 0.573 | Improved |
| pretrain on SarSim0 instead of KernelSynth data | NBEATS · M-Series | MASE: 1.033 -> 0.869 | Improved |
Closest papers by meaning
- TempoPFN: Synthetic Pre-training of Linear RNNs for Zero-shot Time Series Forecasting (NeurIPS 2026 · similarity 0.88)
- Zero-shot forecasting of chaotic systems (ICLR 2025 · similarity 0.88)
- TiRex: Zero-Shot Forecasting Across Long and Short Horizons with Enhanced In-Context Learning (NeurIPS 2025 · similarity 0.87)
- TimePFN: Effective Multivariate Time Series Forecasting with Synthetic Data (AAAI 2025 · similarity 0.86)
- Tiny Time Mixers (TTMs): Fast Pre-trained Models for Enhanced Zero/Few-Shot Forecasting of Multivariate Time Series (NeurIPS 2024 · similarity 0.86)
- Cherry-Picking in Time Series Forecasting: How to Select Datasets to Make Your Model Shine (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.
