Non-Stationary long-term time series prediction based on encoding improvements and frequency domain enhancement

WANG Jian-xiao, SHEN Shi-kai, SHE Yu-mei, YANG Bin, HONG Yi, TAO Yu-hu

PDF(1285 KB)
PDF(1285 KB)
Journal of Yunnan University of Nationalities(Natural Sciences Edition) ›› 2025, Vol. 34 ›› Issue (03) : 350-355. DOI: 10.3969/j.issn.1672-8513.2025.03.013

Non-Stationary long-term time series prediction based on encoding improvements and frequency domain enhancement

Author information +
History +

Abstract

To address the issues in the Informer model， which does not account for the non - stationarity and frequency domain information in real - world data， a non - stationary long-term time series prediction model is proposed. The core idea involves encoding improvements and frequency domain enhancement. To restore non - stationary information to temporal dependencies， the model uses the time absolute position encoding to extract interdependencies between time points. Additionally， the frequency domain enhancement with channel attention， based on the discrete cosine transform， adaptively captures the interdependencies between channels in the frequency domain， thereby improving predictability. Experimental results show that， compared to other models， the proposed model achieves an average reduction of 58.4% in mean squared error （MSE） on the dataset， with a maximum reduction of 66.5%.

Key words

Long - term time series prediction / time absolute position encoding / frequency enhanced channel attention / discrete cosine transform

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations
WANG Jian-xiao , SHEN Shi-kai , SHE Yu-mei , et al . Non-Stationary long-term time series prediction based on encoding improvements and frequency domain enhancement. Journal of Yunnan University of Nationalities(Natural Sciences Edition). 2025, 34(03): 350-355 https://doi.org/10.3969/j.issn.1672-8513.2025.03.013

References

List( Publishing order | Descend order by publishing year | Descend order by cited within ) Chart analysis
1
YE H， CHEN J， GONG S， et al. ATFNet： adaptive time - frequency ensembled network for long-term time series forecasting［EB/OL］. 2024 - 04 - 08/2025 - 04 - 30.
https://arxiv.org/abs/2404.05192
本文引用 [1]
2
ZHOU H， ZHANG S， PENG J， et al. Informer： beyond efficient transformer for long sequence time-series forecasting［C］//Proceedings of the AAAI Conference on Artificial Intelligence， 2021， 35（12）： 11106 - 11115.
本文引用 [2]
3
OGASAWARA E， MARTINEZ L C， OLIVEIRA D， et al. Adaptive normalization： a novel data normalization approach for non - stationary time series［C］//The 2010 International Joint Conference on Neural Networks （IJCNN）. IEEE， 2010： 1 - 8.
本文引用 [1]
4
PASSALIS N， TEFAS A， KANNIAINEN J， et al. Deep adaptive input normalization for time series forecasting［J］. IEEE Transactions on Neural Networks and Learning Systems， 2019， 31（9）： 3760 - 3765.
5
KIM T， KIM J， TAE Y， et al. Reversible instance normalization for accurate time - series forecasting against distribution shift［C］//ICLR， 2022.
本文引用 [1]
6
LIU Y， WU H， WANG J， et al. Non-stationary transformers： exploring the stationarity in time series forecasting［J］. Advances in Neural Information Processing Systems， 2022， 35： 9881 - 9893.
本文引用 [2]
7
ZENG A， CHEN M， ZHANG L， et al. Are transformers effective for time series forecasting？［C］//Proceedings of the AAAI Conference on Artificial Intelligence， 2023， 37（9）： 11121 - 11128.
本文引用 [1]
8
WU H， XU J， WANG J， et al. Autoformer： decomposition transformers with auto-correlation for long-term series forecasting［C］//Advances in Neural Information Processing Systems， 2021， 34： 22419 - 22430.
本文引用 [1]
9
WU H， XU J， WANG J， et al. TimesNet： temporal 2D-variation modeling for general time series analysis ［J］. arXiv preprint arXiv：2210.02186， 2022.
本文引用 [1]
10
GIBBS J W. Fourier's series［J］. Nature， 1899， 59（1539）： 606.
本文引用 [2]
11
VASWANI A， SHAZEER N， PARMAR N， et al. Attention is all you need［C］//Advances in Neural Information Processing Systems， 2017： 30.
本文引用 [2]
12
DEVLIN J， CHANG M W， LEE K， et al. BERT： pre training of deep bidirectional transformers for language understanding［C］//Proceedings of the 2019 conference of the North American chapter of the Association for Computational Linguistics： Human Language Technologies， 2019： 4171 - 4186.
本文引用 [1]
13
HU J， SHEN L， ALBANIE S， et al. Squeeze - and - excitation networks［C］//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition， 2018： 7132 - 7141.
本文引用 [1]
14
JIANG M， ZENG P， WANG K， et al. FECAM： frequency enhanced channel attention mechanism for time series forecasting［J］. Advanced Engineering Informatics， 2023， 58： 102158.
本文引用 [1]
15
CHAVES S S， LYNFYIELD R， LINDEGREN M L， et al. The US influenza hospitalization surveillance network［J］. Emerging Infectious Diseases， 2015， 21（9）： 1543.
本文引用 [2]
16
KITAEV N， KAISER L， LEVSKAYA A， et al. Reformer： the efficient transformer［J］. arXiv preprint arXiv：2001.04451， 2020： 1 - 12.
17
LAI G， CHANG W， YANG Y， et al. Modeling long- and short-term temporal patterns with deep neural networks［C］//The 41st international ACM SIGIR Conference on Research & Development in Information Retrieval. 2018： 95 - 104.
本文引用 [1]
18
LI S， JIN X， XUAN Y， et al. Enhancing the locality and breaking the memory bottleneck of transformer on time series forecasting［J］. Advances in Neural Information Processing Systems， 2019， 32： 5243 - 5253.

Comments

PDF(1285 KB)

Accesses

Citation

Detail
Sections
Recommended
Copyright © Journal of Yunnan University of Nationalities(Natural Sciences Edition), All Rights Reserved.
Powered by Beijing Magtech Co. Ltd.

/

〈 〉