Research map / Time series / Causal and generative time series
WaveBound: Dynamic Error Bounds for Stable Time Series Forecasting
Youngin Cho, Daejin Kim, Dongmin Kim, Mohammad Azam Khan, Jaegul Choo
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 WaveBound regularization | Autoformer · ETTm2 multivariate M=96 | MSE: 0.262 -> 0.204 | Improved |
| add WaveBound regularization | LSTNet · ETTm2 multivariate M=96 | MSE: 0.455 -> 0.268 | Improved |
| add WaveBound regularization | LSTNet · ETTm2 multivariate M=96 | MAE: 0.511 -> 0.368 | Improved |
| add WaveBound regularization | Autoformer · ETTm2 multivariate M=720 | MSE: 0.446 -> 0.413 | Improved |
| add WaveBound regularization | N-BEATS · ETTm2 univariate M=96 | MSE: 0.073 -> 0.067 | Improved |
| add WaveBound regularization | Informer · ECL univariate M=720 | MSE: 0.631 -> 0.378 | Improved |
| add WaveBound regularization | Informer · Synthetic sigma=0.5 | MSE: 0.605 -> 0.281 | Improved |
| replace fixed-constant flooding with WaveBound (Indiv.) bounds (ablation) | Autoformer · ECL M=96 | MSE: 0.202 -> 0.176 | Improved |
| use original flooding with a constant flood level (ablation) | Autoformer · ECL M=96 | MSE: 0.202 -> 0.194 | Improved |
| use original flooding with a constant flood level (ablation) | Informer · ECL M=96 | MSE: 0.335 -> 0.335 | No clear effect |
| use constant flooding per feature and time step (ablation) | Pyraformer · ECL M=96 | MSE: 0.256 -> 0.257 | No clear effect |
| bound error individually per feature and time step instead of averaged (ablation) | Informer · ECL M=96 | MSE: 0.302 -> 0.289 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Informer — Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting (AAAI 2021) · better on 5 of 5 shared results
- Autoformer — Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting (NeurIPS 2021) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Loss Shaping Constraints for Long-Term Time Series Forecasting (ICML 2024 · similarity 0.89)
- Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization (ICML 2025 · similarity 0.88)
- SynTSBench: Rethinking Temporal Pattern Learning in Deep Learning Models for Time Series (NeurIPS 2025 · similarity 0.87)
- DDN: Dual-domain Dynamic Normalization for Non-stationary Time Series Forecasting (NeurIPS 2024 · similarity 0.87)
- Abstain Mask Retain Core: Time Series Prediction by Adaptive Masking Loss with Representation Consistency (NeurIPS 2025 · similarity 0.86)
- Scaling Law for Time Series Forecasting (NeurIPS 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.
