Research map / Time series / Long-term forecasting
Transparent Networks for Multivariate Time Series
Minkyu Kim, Suan Lee, Jinho Kim
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 masked multi-head attention and positional encoding (temporal module) (ablation) | GATSM · LSST | accuracy: 0.392 -> 0.570 | Improved |
| add masked multi-head attention and positional encoding (temporal module) (ablation) | GATSM · NATOPS | accuracy: 0.192 -> 0.956 | Improved |
| add positional encoding alone without MHA (ablation) | GATSM base · Rainfall / AirQuality / NATOPS | R2 / accuracy: 0.299 -> 0.299 (AirQuality) | No clear effect |
| add positional encoding to MHA (ablation) | GATSM · NATOPS | accuracy: 0.856 -> 0.956 | Improved |
| add temporal module (ablation) | GATSM · Mortality / Sepsis | AUROC: 0.854 -> 0.853 (Mortality) | No clear effect |
| use NBM instead of linear feature function (ablation) | GATSM · Energy | R2: 0.283 -> 0.493 | Improved |
| use NBM instead of linear feature function (ablation) | GATSM · NATOPS | accuracy: 0.972 -> 0.956 | Hurt |
| use sinusoidal instead of learnable positional encoding (ablation) | GATSM · 8 datasets | R2 / AUROC / accuracy: 0.519 -> 0.493 (Energy) | No clear effect |
| increase number of basis functions beyond 100 (ablation) | GATSM · Energy / Heartbeat | R2 / AUROC | No clear effect |
| use GATSM instead of Transformer | GATSM · Energy | R2: 0.263 -> 0.493 | Improved |
| use GATSM instead of Transformer | GATSM · NATOPS | accuracy: 0.967 -> 0.956 | Hurt |
| use GATSM instead of NATM | GATSM · Heartbeat | AUROC: 0.724 -> 0.843 | Improved |
Closest papers by meaning
- T-Rep: Representation Learning for Time Series using Time-Embeddings (ICLR 2024 · similarity 0.86)
- SynTSBench: Rethinking Temporal Pattern Learning in Deep Learning Models for Time Series (NeurIPS 2025 · similarity 0.85)
- TimeMixer++: A General Time Series Pattern Machine for Universal Predictive Analysis (ICLR 2025 · similarity 0.85)
- Explaining Time Series via Contrastive and Locally Sparse Perturbations (ICLR 2024 · similarity 0.85)
- Towards a General Time Series Forecasting Model with Unified Representation and Adaptive Transfer (ICML 2025 · similarity 0.84)
- Explain Temporal Black-Box Models via Functional Decomposition (ICML 2024 · similarity 0.84)
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.
