FreDN: Spectral Disentanglement for Time Series Forecasting via Learnable Frequency Decomposition
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (24) · 19623-19631
DOI: 10.1609/aaai.v40i24.39042 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Zhongde An und 6 weiteren Autor:innen. Die Arbeit wurde bisher 9-mal zitiert (Crossref).
Abstract (Originalsprache)
Time series forecasting is essential in a wide range of real world applications. Recently, frequency-domain methods have attracted increasing interest for their ability to capture global dependencies. However, when applied to non-stationary time series, these methods encounter the spectral entanglement and the computational burden of complex-valued learning. The spectral entanglement refers to the overlap of trends, periodicities, and noise across the spectrum due to spectral leakage and the presence of non-stationarity. However, existing decompositions are not suited to resolving spectral entanglement. To address this, we propose the Frequency Decomposition Network (FreDN), which introduces a learnable Frequency Disentangler module to separate trend and periodic components directly in the frequency domain. Furthermore, we propose a theoretically supported ReIm Block to reduce the complexity of complex-valued operations while maintaining performance. We also re-examine the frequency-domain loss function and provide new theoretical insights into its effectiveness. Extensive experiments on seven long-term forecasting benchmarks demonstrate that FreDN outperforms state-of-the-art methods by up to 10%. Furthermore, compared with standard complex-valued architectures, our real-imaginary shared-parameter design reduces the parameter count and computational cost by at least 50%.
Zitieren
APA
An, Z., You, J., Li, J., Tang, Y., Li, W., Du, H., & Du, S. (2026). FreDN: Spectral Disentanglement for Time Series Forecasting via Learnable Frequency Decomposition. Proceedings of the AAAI Conference on Artificial Intelligence, 40(24), 19623-19631. https://doi.org/10.1609/aaai.v40i24.39042
Vancouver
An Z, You J, Li J, Tang Y, Li W, Du H, et al. FreDN: Spectral Disentanglement for Time Series Forecasting via Learnable Frequency Decomposition. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(24):19623-19631. doi:10.1609/aaai.v40i24.39042
BibTeX
@article{An2026,
author = {An, Zhongde and You, Jinhong and Li, Jiyanglin and Tang, Yiming and Li, Wen and Du, Heming and Du, Shouguo},
title = {FreDN: Spectral Disentanglement for Time Series Forecasting via Learnable Frequency Decomposition},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {24},
pages = {19623--19631},
doi = {10.1609/aaai.v40i24.39042}
}
EUVIMED