Research map / Time series / Long-term forecasting
FourierGNN: Rethinking Multivariate Time Series Forecasting from a Pure Graph Perspective
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 FourierGNN instead of the strongest reported baseline | MTS forecasting · Solar | MAE: 0.123 -> 0.120 | Helped |
| use FourierGNN instead of the strongest reported baseline | MTS forecasting · Wiki | MAE: 0.044 -> 0.041 | Helped |
| use FourierGNN instead of the strongest reported baseline | MTS forecasting · Traffic | MAE: 0.013 -> 0.011 | Helped |
| use FourierGNN instead of the strongest reported baseline | MTS forecasting · ECG | MAE: 0.055 -> 0.052 | Helped |
| use FourierGNN instead of the strongest reported baseline | MTS forecasting · Electricity | MAE: 0.055 -> 0.051 | Helped |
| use FourierGNN instead of the strongest reported baseline | MTS forecasting · COVID-19 | MAE: 0.152 -> 0.123 | Helped |
| add node embedding (ablation) | FourierGNN · METR-LA | MAE: 0.053 -> 0.050 | Helped |
| use different FGOs across diffusion steps (ablation) | FourierGNN · METR-LA | MAE: 0.055 -> 0.050 | Helped |
| add the K=0 residual layer (ablation) | FourierGNN · METR-LA | MAE: 0.054 -> 0.050 | Helped |
| sum outputs across diffusion orders (ablation) | FourierGNN · METR-LA | MAE: 0.054 -> 0.050 | Helped |
| use n-invariant rather than n-variant FGO (ablation) | FourierGNN · ECG | MAE: 0.053 -> 0.052 | Helped |
| use n-invariant rather than n-variant FGO (ablation) | FourierGNN · COVID-19 | MAE: 0.129 -> 0.123 | Helped |
Closest papers by meaning
- Irregular Multivariate Time Series Forecasting: A Transformable Patching Graph Neural Networks Approach (ICML 2024 · similarity 0.89)
- HyperIMTS: Hypergraph Neural Network for Irregular Multivariate Time Series Forecasting (ICML 2025 · similarity 0.89)
- Fully-Connected Spatial-Temporal Graph for Multivariate Time-Series Data (AAAI 2024 · similarity 0.89)
- Graph-based Forecasting with Missing Data through Spatiotemporal Downsampling (ICML 2024 · similarity 0.89)
- Rethinking Fourier Transform from A Basis Functions Perspective for Long-term Time Series Forecasting (NeurIPS 2024 · similarity 0.88)
- TimePFN: Effective Multivariate Time Series Forecasting with Synthetic Data (AAAI 2025 · 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.