Research map / Time series / Long-term forecasting
Pay Attention to Evolution: Time Series Forecasting With Deep Graph-Evolution Learning
Gabriel Spadon, Shenda Hong, Bruno Brandoli Machado, Stan Matwin, Jose Fernando Rodrigues Junior, Jimeng Sun
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 GSE layers (full ReGENN vs best no-GSE ablation) (ablation) | ReGENN · SARS-CoV-2 | MAE, RMSE, MSLE: 20.81% MAE, 19.77% RMSE, 35.72% MSLE | Improved |
| add GSE layers over best no-GSE ablation (ablation) | ReGENN · Brazilian Weather | MAE, RMSE: 8.42% on the MAE and 2.16% on the RMSE | Improved |
| add GSE layers over bidirectional LSTM decoder (ablation) | ReGENN · PhysioNet 2012 | MAE, RMSE: 1.05% on the MAE and 0.98% on the RMSE | Improved |
| add Autoregression component (ablation) | ReGENN · SARS-CoV-2 | MAE: 8208.53 -> 165.41 | Improved |
| add Autoregression component (ablation) | ReGENN · Brazilian Weather | MAE: 2.00 -> 1.92 | Improved |
| add Autoregression component (ablation) | ReGENN · PhysioNet 2012 | RMSE: 49.72 -> 47.31 | Improved |
| add variable-axis decoder (ablation) | ReGENN · SARS-CoV-2 | MSLE: 0.15 -> 0.05 | Improved |
| add variable-axis decoder (ablation) | ReGENN · Brazilian Weather | MSLE: 0.27 -> 0.28 | No clear effect |
| add variable-axis decoder (ablation) | ReGENN · PhysioNet 2012 | MAE: 18.23 -> 18.22 | No clear effect |
| combine network with Autoregression instead of Autoregression only (ablation) | ReGENN · SARS-CoV-2 | MAE: 11529.79 -> 165.41 | Improved |
| use ReGENN instead of ensemble baselines (second-best Exponential Smoothing) | ReGENN · SARS-CoV-2 (120 days) | MAE, RMSE, MSLE: 75.21% on the MAE, 64.87% on the RMSE, and 79.61% on the MSLE | Improved |
| use ReGENN instead of Elman RNN | ReGENN · Brazilian Weather | MAE, RMSE, MSLE: 11.95% on the MAE, 11.96% on the RMSE, and 25.84% on the MSLE | Improved |
Closest papers by meaning
- Irregular Multivariate Time Series Forecasting: A Transformable Patching Graph Neural Networks Approach (ICML 2024 · similarity 0.88)
- SynTSBench: Rethinking Temporal Pattern Learning in Deep Learning Models for Time Series (NeurIPS 2025 · similarity 0.87)
- TimeES: Probabilistic and Deterministic Time Series Forecasting via Evolutionary Spectra (NeurIPS 2026 · similarity 0.86)
- TStruct: Learning Shared Temporal Structures for Long-Term Time-Series Forecasting (NeurIPS 2026 · similarity 0.86)
- Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Experts (ICLR 2025 · similarity 0.86)
- TimePFN: Effective Multivariate Time Series Forecasting with Synthetic Data (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.
