Research map / Time series / Long-term forecasting
Deep Factors for Forecasting
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 global deep factors with local probabilistic effects instead of the strongest reported baseline | DF-RNN versus strongest baseline · Electricity, 72-hour | P50QL: 0.149 -> 0.101 | Helped |
| Use global deep factors with local probabilistic effects instead of the strongest reported baseline | DF-RNN versus strongest baseline · NYC taxi, 72-hour | P50QL: 0.397 -> 0.212 | Helped |
| Use global deep factors with local probabilistic effects instead of the strongest reported baseline | DF-RNN versus strongest baseline · Traffic, 72-hour | P50QL: 0.337 -> 0.190 | Helped |
| Use global deep factors with local probabilistic effects instead of the strongest reported baseline | DF-RNN versus strongest baseline · Uber, 72-hour | P50QL: 0.417 -> 0.344 | Helped |
| Use global deep factors with local probabilistic effects instead of the strongest reported baseline | DF-RNN versus strongest baseline · Electricity, 24-hour | P50QL: 0.124 -> 0.112 | Helped |
| Use global deep factors with local probabilistic effects instead of the strongest reported baseline | DF-RNN versus strongest baseline · NYC taxi, 24-hour | P50QL: 0.328 -> 0.239 | Helped |
| Use global deep factors with local probabilistic effects instead of the strongest reported baseline | DF-RNN versus strongest baseline · Traffic, 24-hour | P50QL: 0.296 -> 0.225 | Helped |
| Use global deep factors with local probabilistic effects instead of the strongest reported baseline | DF-RNN versus strongest baseline · Uber, 24-hour | P50QL: 0.353 -> 0.425 | Hurt |
| Use global deep factors with local probabilistic effects instead of the strongest reported baseline | DF-RNN versus strongest baseline · Traffic, 24-hour | P90QL: 0.149 -> 0.159 | Hurt |
| Replace the DF-RNN local noise process with a level-trend LDS (ablation) | DF-RNN · Electricity, horizon 24 | P50QL: 0.144 -> 0.153 | Hurt |
| Replace the DF-RNN local noise process with a level-trend LDS (ablation) | DF-RNN · Electricity, horizon 24 | P90QL: 0.101 -> 0.135 | Hurt |
| Replace the DF-RNN local noise process with a level-trend LDS (ablation) | DF-RNN · Electricity, horizon 24 | RMSE: 888.285 -> 1012.551 | Hurt |
Closest papers by meaning
- Multivariate Probabilistic Time Series Forecasting with Correlated Errors (NeurIPS 2024 · similarity 0.88)
- Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Experts (ICLR 2025 · similarity 0.87)
- ProbTS: Benchmarking Point and Distributional Forecasting across Diverse Prediction Horizons (NeurIPS 2024 · similarity 0.87)
- LETS Forecast: Learning Embedology for Time Series Forecasting (ICML 2025 · similarity 0.87)
- Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization (ICML 2025 · similarity 0.87)
- SIN: Selective and Interpretable Normalization for Long-Term Time Series Forecasting (ICML 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.