Research map / Time series / Long-term forecasting
FIPN: Forward Self-Organizing Interpretable Polynomial Networks for Time Series Forecasting
YiZhen Wang, Zheng Wang, EUN-HU KIM, Zunwei Fu
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 Fourier-enhanced consequents (ablation) | FIPN · 8 benchmarks (avg) | MSE: 0.3492 -> 0.3310 | Improved |
| add Fourier-enhanced consequents (ablation) | FIPN · 8 benchmarks (avg) | MAE: 0.3566 -> 0.3419 | Improved |
| add redundancy-controlled selection and dropout (ablation) | FIPN · 8 benchmarks (avg) | MSE: 0.3399 -> 0.3310 | Improved |
| add redundancy-controlled selection and dropout (ablation) | FIPN · 8 benchmarks (avg) | MAE: 0.3494 -> 0.3419 | Improved |
| add fuzzy rule gating (ablation) | FIPN · 8 benchmarks (avg) | MSE: 0.3346 -> 0.3310 | Improved |
| add fuzzy rule gating (ablation) | FIPN · 8 benchmarks (avg) | MAE: 0.3453 -> 0.3419 | Improved |
| add l2 regularization to closed-form fitting (ablation) | FIPN · Traffic (H=96) | MSE: 0.403 -> 0.392 | Improved |
| add l2 regularization to closed-form fitting (ablation) | FIPN · Electricity (H=96) | MSE: 0.141 -> 0.137 | Improved |
| increase number of fuzzy rules C beyond a moderate value (ablation) | FIPN · ETTh1, ETTm1, Traffic, Exchange | MSE | Mixed |
| vary fuzzification coefficient gamma (ablation) | FIPN · ETTh1, ETTm1, Traffic, Exchange | MSE | Mixed |
| use FIPN instead of Crossformer-scale models for parameter count | FIPN · ETTh1 (H=720) | parameters: 17.40M -> 0.89K | Improved |
| use FIPN versus TimeStacker | FIPN · Traffic (avg) | MSE: 0.508 -> 0.441 | 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.
- TimeStacker — TimeStacker: A Novel Framework with Multilevel Observation for Capturing Nonstationary Patterns in Time Series Forecasting (ICML 2025) · better on 6 of 8 shared results
- iTransformer — iTransformer: Inverted Transformers Are Effective for Time Series Forecasting (ICLR 2024) · better on 7 of 8 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
- PGN: The RNN's New Successor is Effective for Long-Range Time Series Forecasting (NeurIPS 2024 · similarity 0.85)
- FilterNet: Harnessing Frequency Filters for Time Series Forecasting (NeurIPS 2024 · similarity 0.84)
- Rethinking Fourier Transform from A Basis Functions Perspective for Long-term Time Series Forecasting (NeurIPS 2024 · similarity 0.84)
- Structured Matrix Basis for Multivariate Time Series Forecasting with Interpretable Dynamics (NeurIPS 2024 · similarity 0.84)
- HyperIMTS: Hypergraph Neural Network for Irregular Multivariate Time Series Forecasting (ICML 2025 · similarity 0.84)
- TimePFN: Effective Multivariate Time Series Forecasting with Synthetic Data (AAAI 2025 · 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.
