Nested Spatio-Temporal Time Series Forecasting
Abstract
Spatiotemporal forecasting is critical for real-world applications like traffic management, yet capturing reliable interactions remains challenging under noisy and non-stationary conditions. Existing methods primarily rely on historical spatial priors, often failing to account for evolving temporal correlations and suffering from 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 employ a spectral clustering-based approach to construct semantically coherent regions, providing both theoretical and empirical evidence that this representation effectively filters systematic noise while preserving essential trends. Building on this, we develop a progressive coarse-to-fine predictor to integrate these representative features into the inference process. This enables the model to leverage trend predictions to anticipate dynamic anomalies, such as periodic offsets, in advance. Furthermore, extensive experiments on multiple high-dimensional datasets demonstrate that our method consistently outperforms state-of-the-art baselines, validating the effectiveness of future macro-guided nested forecasting.
Cite
@article{arxiv.2605.16447,
title = {Nested Spatio-Temporal Time Series Forecasting},
author = {Yinghao Ai and Yukai Zhou and Ruoxi Jiang and Junyi An and Chao Qu and Zhijian Zhou and Shiyu Wang and Fenglei Cao and Zenglin Xu and Furao Shen and Yuan Qi},
journal= {arXiv preprint arXiv:2605.16447},
year = {2026}
}
Comments
Accept by ICML 2026