Nested Spatio-Temporal Time Series Forecasting

Yinghao Ai, Yukai Zhou, Ruoxi Jiang, Junyi An, Chao Qu, Zhijian Zhou, Shiyu Wang, Fenglei Cao, Zenglin Xu, Furao Shen, Yuan Qi
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:1384-1403, 2026.

Abstract

Spatio-temporal forecasting is critical for real-world applications like traffic management, yet capturing complex interactions under high-noise conditions remains challenging. While current methods have shown improved accuracy using spatial physical priors, they often struggle with evolving temporal correlations and systematic errors. In this work, we propose a nested forecasting framework that couples future macro-level regional trends with micro-level historical observations, enabling top-down guidance from abstract future representations for fine-grained forecasting. Specifically, we construct semantically coherent regions via spectral clustering and design a progressive coarse-to-fine predictor to inject macro-dynamics into node-level forecasting. Extensive experiments on multiple real-world datasets demonstrate that our method consistently outperforms state-of-the-art baselines, validating the effectiveness of future macro-guided nested forecasting.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-ai26f, title = {Nested Spatio-Temporal Time Series Forecasting}, author = {Ai, Yinghao and Zhou, Yukai and Jiang, Ruoxi and An, Junyi and Qu, Chao and Zhou, Zhijian and Wang, Shiyu and Cao, Fenglei and Xu, Zenglin and Shen, Furao and Qi, Yuan}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {1384--1403}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/ai26f/ai26f.pdf}, url = {https://proceedings.mlr.press/v306/ai26f.html}, abstract = {Spatio-temporal forecasting is critical for real-world applications like traffic management, yet capturing complex interactions under high-noise conditions remains challenging. While current methods have shown improved accuracy using spatial physical priors, they often struggle with evolving temporal correlations and systematic errors. In this work, we propose a nested forecasting framework that couples future macro-level regional trends with micro-level historical observations, enabling top-down guidance from abstract future representations for fine-grained forecasting. Specifically, we construct semantically coherent regions via spectral clustering and design a progressive coarse-to-fine predictor to inject macro-dynamics into node-level forecasting. Extensive experiments on multiple real-world datasets demonstrate that our method consistently outperforms state-of-the-art baselines, validating the effectiveness of future macro-guided nested forecasting.} }
Endnote
%0 Conference Paper %T Nested Spatio-Temporal Time Series Forecasting %A Yinghao Ai %A Yukai Zhou %A Ruoxi Jiang %A Junyi An %A Chao Qu %A Zhijian Zhou %A Shiyu Wang %A Fenglei Cao %A Zenglin Xu %A Furao Shen %A Yuan Qi %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-ai26f %I PMLR %P 1384--1403 %U https://proceedings.mlr.press/v306/ai26f.html %V 306 %X Spatio-temporal forecasting is critical for real-world applications like traffic management, yet capturing complex interactions under high-noise conditions remains challenging. While current methods have shown improved accuracy using spatial physical priors, they often struggle with evolving temporal correlations and systematic errors. In this work, we propose a nested forecasting framework that couples future macro-level regional trends with micro-level historical observations, enabling top-down guidance from abstract future representations for fine-grained forecasting. Specifically, we construct semantically coherent regions via spectral clustering and design a progressive coarse-to-fine predictor to inject macro-dynamics into node-level forecasting. Extensive experiments on multiple real-world datasets demonstrate that our method consistently outperforms state-of-the-art baselines, validating the effectiveness of future macro-guided nested forecasting.
APA
Ai, Y., Zhou, Y., Jiang, R., An, J., Qu, C., Zhou, Z., Wang, S., Cao, F., Xu, Z., Shen, F. & Qi, Y.. (2026). Nested Spatio-Temporal Time Series Forecasting. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:1384-1403 Available from https://proceedings.mlr.press/v306/ai26f.html.

Related Material