Research map / Time series / Long-term forecasting
FACT: Fine-grained Across-variable Convolution for Multivariate Time Series Forecasting
Huiqiang Wang, Jieming Shi, Li Qing
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 time-domain modeling alongside frequency (ablation) | FACT · METR-LA | MSE: 0.722 -> 0.704 | Improved |
| add frequency-domain modeling alongside time (ablation) | FACT · METR-LA | MSE: 0.755 -> 0.704 | Improved |
| add frequency-domain modeling alongside time (ablation) | FACT · Traffic2 | MSE: 0.215 -> 0.204 | Improved |
| add time-domain modeling alongside frequency (ablation) | FACT · Traffic2 | MSE: 0.212 -> 0.204 | Improved |
| add frequency-domain modeling alongside time (ablation) | FACT · PEMS03 | MSE: 0.089 -> 0.086 | Improved |
| add time-domain modeling alongside frequency (ablation) | FACT · PEMS03 | MSE: 0.087 -> 0.086 | Improved |
| add time-domain modeling (ablation) | FACT · Traffic | MSE: 0.426 -> 0.425 | No clear effect |
| add time-domain modeling (ablation) | FACT · Solar | MSE: 0.198 -> 0.197 | No clear effect |
| add 2D reshaping of variables instead of 1D convolution (ablation) | FACT · FACT | MSE | Improved |
| add multi-dilated convolution versus no dilation (ablation) | FACT · FACT | MSE | Improved |
| use Inception backbone instead of standard 2D convolution (ablation) | FACT · FACT | parameters (M): 21.59 -> 21.94 | Improved |
| increase layers from 3 to 7 (ablation) | FACT · FACT | MSE: 0.210 -> 0.201 | Mixed |
Closest papers by meaning
- MSGNet: Learning Multi-Scale Inter-series Correlations for Multivariate Time Series Forecasting (AAAI 2024 · similarity 0.87)
- TimeXer: Empowering Transformers for Time Series Forecasting with Exogenous Variables (NeurIPS 2024 · similarity 0.87)
- Multivariate Probabilistic Time Series Forecasting with Correlated Errors (NeurIPS 2024 · similarity 0.86)
- Revitalizing Multivariate Time Series Forecasting: Learnable Decomposition with Inter-Series Dependencies and Intra-Series Variations Modeling (ICML 2024 · similarity 0.86)
- HDT: Hierarchical Discrete Transformer for Multivariate Time Series Forecasting (AAAI 2025 · similarity 0.86)
- CATS: Enhancing Multivariate Time Series Forecasting by Constructing Auxiliary Time Series as Exogenous Variables (ICML 2024 · 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.
