Research map / Time series / Long-term forecasting
IdealTSF: Can Non-Ideal Data Contribute to Enhancing the Performance of Time Series Forecasting Models?
Hua Wang, Jinghao Lu, Fan Zhang
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 negative sample pre-training (ablation) | IdealTSF · ETTh2 | MSE: 0.420 -> 0.338 | Improved |
| add negative sample pre-training (ablation) | IdealTSF · ETTm1 | MSE: 0.465 -> 0.409 | Improved |
| add negative sample pre-training (ablation) | IdealTSF · ETTm2 | MSE: 0.301 -> 0.247 | Improved |
| add negative sample pre-training (ablation) | IdealTSF · ETTh1 | MSE: 0.436 -> 0.402 | Improved |
| add positive sample generation (ablation) | IdealTSF · ETTh2 | MSE: 0.347 -> 0.338 | Improved |
| add positive sample generation (ablation) | IdealTSF · ETTm1 | MSE: 0.417 -> 0.409 | Improved |
| add EcoSystem Optimizer (ECOS) (ablation) | IdealTSF · ETTh2 | MSE: 0.388 -> 0.338 | Improved |
| add EcoSystem Optimizer (ECOS) (ablation) | IdealTSF · ETTm2 | MSE: 0.270 -> 0.247 | Improved |
| add ECOS on top of negative-sample-free model (ablation) | IdealTSF · ETTm1 | MSE: 0.468 -> 0.436 | Improved |
| use IdealTSF instead of Twinformer / TimeMixer++ | IdealTSF · PEMS07 and ETTm1 | MSE: PEMS07 0.118 vs Twinformer 0.094; ETTm1 0.409 vs TimeMixer++ 0.369 | Hurt |
| add ECOS optimizer to Adam or SGD (ablation) | CIFAR-10 classifier · CIFAR-10 | accuracy (%): 80 -> 90 (Adam); 80 -> nearly 90 (SGD) | Improved |
| use ECOS instead of SGD/RMSprop under PGD attack (ablation) | IdealTSF · adversarial robustness (FGSM/PGD) | test accuracy (%) | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- TimeMixer++ — TimeMixer: Decomposable Multiscale Mixing for Time Series Forecasting (ICLR 2024) · better on 6 of 7 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
- SynTSBench: Rethinking Temporal Pattern Learning in Deep Learning Models for Time Series (NeurIPS 2025 · similarity 0.84)
- RobustTSF: Towards Theory and Design of Robust Time Series Forecasting with Anomalies (ICLR 2024 · similarity 0.84)
- Selective Learning for Deep Time Series Forecasting (NeurIPS 2025 · similarity 0.84)
- Sequence Complementor: Complementing Transformers for Time Series Forecasting with Learnable Sequences (AAAI 2025 · similarity 0.84)
- ProbTS: Benchmarking Point and Distributional Forecasting across Diverse Prediction Horizons (NeurIPS 2024 · similarity 0.83)
- SIN: Selective and Interpretable Normalization for Long-Term Time Series Forecasting (ICML 2024 · similarity 0.83)
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.
