FreqMoE: Enhancing Time Series Forecasting through Frequency Decomposition Mixture of Experts

Ziqi Liu
Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, PMLR 258:3430-3438, 2025.

Abstract

Long-term time series forecasting is essential in areas like finance and weather prediction. Besides traditional methods that operate in the time domain, many recent models transform time series data into the frequency domain to better capture complex patterns. However, these methods often use filtering techniques to remove certain frequency signals as noise, which may unintentionally discard important information and reduce prediction accuracy. To address this, we propose the Frequency Decomposition Mixture-of-Experts (FreqMoE) model, which dynamically decomposes time series data into frequency bands, each processed by a specialized expert. A gating mechanism adjusts the importance of each output of expert based on frequency characteristics, and the aggregated results are fed into a prediction module that iteratively refines the forecast using residual connections. Our experiments demonstrate that FreqMoE outperforms state-of-the-art models, achieving the best performance on 51 out of 70 metrics across all tested datasets, while significantly reducing the number of required parameters to under 50k, providing notable efficiency advantages. Code is available at: \url{https://github.com/sunbus100/FreqMoE-main}

Cite this Paper

BibTeX
@InProceedings{pmlr-v258-liu25i, title = {FreqMoE: Enhancing Time Series Forecasting through Frequency Decomposition Mixture of Experts}, author = {Liu, Ziqi}, booktitle = {Proceedings of The 28th International Conference on Artificial Intelligence and Statistics}, pages = {3430--3438}, year = {2025}, editor = {Li, Yingzhen and Mandt, Stephan and Agrawal, Shipra and Khan, Emtiyaz}, volume = {258}, series = {Proceedings of Machine Learning Research}, month = {03--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v258/main/assets/liu25i/liu25i.pdf}, url = {https://proceedings.mlr.press/v258/liu25i.html}, abstract = {Long-term time series forecasting is essential in areas like finance and weather prediction. Besides traditional methods that operate in the time domain, many recent models transform time series data into the frequency domain to better capture complex patterns. However, these methods often use filtering techniques to remove certain frequency signals as noise, which may unintentionally discard important information and reduce prediction accuracy. To address this, we propose the Frequency Decomposition Mixture-of-Experts (FreqMoE) model, which dynamically decomposes time series data into frequency bands, each processed by a specialized expert. A gating mechanism adjusts the importance of each output of expert based on frequency characteristics, and the aggregated results are fed into a prediction module that iteratively refines the forecast using residual connections. Our experiments demonstrate that FreqMoE outperforms state-of-the-art models, achieving the best performance on 51 out of 70 metrics across all tested datasets, while significantly reducing the number of required parameters to under 50k, providing notable efficiency advantages. Code is available at: \url{https://github.com/sunbus100/FreqMoE-main}} }
Endnote
%0 Conference Paper %T FreqMoE: Enhancing Time Series Forecasting through Frequency Decomposition Mixture of Experts %A Ziqi Liu %B Proceedings of The 28th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2025 %E Yingzhen Li %E Stephan Mandt %E Shipra Agrawal %E Emtiyaz Khan %F pmlr-v258-liu25i %I PMLR %P 3430--3438 %U https://proceedings.mlr.press/v258/liu25i.html %V 258 %X Long-term time series forecasting is essential in areas like finance and weather prediction. Besides traditional methods that operate in the time domain, many recent models transform time series data into the frequency domain to better capture complex patterns. However, these methods often use filtering techniques to remove certain frequency signals as noise, which may unintentionally discard important information and reduce prediction accuracy. To address this, we propose the Frequency Decomposition Mixture-of-Experts (FreqMoE) model, which dynamically decomposes time series data into frequency bands, each processed by a specialized expert. A gating mechanism adjusts the importance of each output of expert based on frequency characteristics, and the aggregated results are fed into a prediction module that iteratively refines the forecast using residual connections. Our experiments demonstrate that FreqMoE outperforms state-of-the-art models, achieving the best performance on 51 out of 70 metrics across all tested datasets, while significantly reducing the number of required parameters to under 50k, providing notable efficiency advantages. Code is available at: \url{https://github.com/sunbus100/FreqMoE-main}
APA
Liu, Z.. (2025). FreqMoE: Enhancing Time Series Forecasting through Frequency Decomposition Mixture of Experts. Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 258:3430-3438 Available from https://proceedings.mlr.press/v258/liu25i.html.

Related Material