Research map / Time series / Long-term forecasting
Delving into Non-Exchangeability for Conformal Prediction in Graph-Structured Multivariate Time Series
Ruichao Guo, Xingyao Han, Luo Wenshui, Zhe Liu, Chen Gong, Hesheng Wang
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 graph-wavelet spectral decomposition (SGWT) to the conformal predictor (ablation) | SCALE · AirQuality (alpha=0.1) | Coverage: 0.9205 (fails) -> 0.9126 (valid) | Improved |
| add graph-wavelet decomposition (SGWT) (ablation) | SCALE · Electricity (alpha=0.05) | PI-Width: 918.76 -> 776.05 | Improved |
| add graph-wavelet decomposition (SGWT) (ablation) | SCALE · METR-LA (alpha=0.1) | PI-Width: 8.88 -> 8.89 | No clear effect |
| add low-frequency conditioning via adaptive gating (ablation) | SCALE · AirQuality (alpha=0.1) | Coverage: 0.1015 -> 0.9126 | Improved |
| add low-frequency conditioning via adaptive gating (ablation) | SCALE · Electricity (alpha=0.05) | Coverage: 0.3946 -> 0.9467 | Improved |
| add low-frequency conditioning via adaptive gating (ablation) | SCALE · METR-LA (alpha=0.1) | Coverage: 0.8968 -> 0.8913 | No clear effect |
| use SCALE instead of CoREL for interval construction | conformal prediction on STGNN residuals · METR-LA (alpha=0.1) | PI-Width: 9.15 -> 8.89 | Improved |
| use SCALE instead of EnbPI | conformal prediction on STGNN residuals · Electricity (alpha=0.05) | PI-Width: 965.59 -> 776.05 | Improved |
| use SCALE instead of NexCP | conformal prediction on STGNN residuals · AirQuality (alpha=0.1) | PI-Width: 40.36 -> 41.22 | No clear effect |
| use SCALE instead of SeqCP | conformal prediction on STGNN residuals · METR-LA (alpha=0.05) | Coverage: 0.9097 -> 0.9436 | Improved |
| use SCALE instead of CoREL | conformal prediction on STGNN residuals · Runtime on METR-LA | runtime (s): 843.9 -> 238.4 | Improved |
| vary the number of SGWT wavelet scales S from 2 to 200 (ablation) | SCALE · METR-LA (alpha=0.1) | Coverage / PI-Width | No clear effect |
Closest papers by meaning
- Conformal Prediction for Time-series Forecasting with Change Points (NeurIPS 2025 · similarity 0.90)
- Relational Conformal Prediction for Correlated Time Series (ICML 2025 · similarity 0.89)
- Filtered Conformal Ellipsoids for Graph-Native Time Series (NeurIPS 2026 · similarity 0.89)
- Copula Conformal prediction for multi-step time series prediction (ICLR 2024 · similarity 0.88)
- Conformal prediction for multi-dimensional time series by ellipsoidal sets (ICML 2024 · similarity 0.87)
- Error-quantified Conformal Inference for Time Series (ICLR 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.
