Select, then Balance: Exploring Exogenous Variable Modeling of Spatio-Temporal Forecasting
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908859577139200 |
|---|---|
| author | Chen, Wei Wu, Yuqian Zhu, Yuanshao Hao, Xixuan Wang, Shiyu Zhou, Xiaofang Liang, Yuxuan |
| author_facet | Chen, Wei Wu, Yuqian Zhu, Yuanshao Hao, Xixuan Wang, Shiyu Zhou, Xiaofang Liang, Yuxuan |
| contents | Spatio-temporal (ST) forecasting is critical for dynamic systems, yet existing methods predominantly rely on modeling a limited set of observed target variables. In this paper, we present the first systematic exploration of exogenous variable modeling for ST forecasting, a topic long overlooked in this field. We identify two core challenges in integrating exogenous variables: the inconsistent effects of distinct variables on the target system and the imbalance effects between historical and future data. To address these, we propose ExoST, a simple yet effective exogenous variable modeling general framework highly compatible with existing ST backbones that follows a "select, then balance" paradigm. Specifically, we design a latent space gated expert module to dynamically select and recompose salient signals from fused exogenous information. Furthermore, a siamese dual-branch backbone architecture captures dynamic patterns from the recomposed past and future representations, integrating them via a context-aware weighting mechanism to ensure dynamic balance. Extensive experiments on real-world datasets demonstrate the ExoST's effectiveness, universality, robustness, and efficiency. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_05779 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Select, then Balance: Exploring Exogenous Variable Modeling of Spatio-Temporal Forecasting Chen, Wei Wu, Yuqian Zhu, Yuanshao Hao, Xixuan Wang, Shiyu Zhou, Xiaofang Liang, Yuxuan Machine Learning Spatio-temporal (ST) forecasting is critical for dynamic systems, yet existing methods predominantly rely on modeling a limited set of observed target variables. In this paper, we present the first systematic exploration of exogenous variable modeling for ST forecasting, a topic long overlooked in this field. We identify two core challenges in integrating exogenous variables: the inconsistent effects of distinct variables on the target system and the imbalance effects between historical and future data. To address these, we propose ExoST, a simple yet effective exogenous variable modeling general framework highly compatible with existing ST backbones that follows a "select, then balance" paradigm. Specifically, we design a latent space gated expert module to dynamically select and recompose salient signals from fused exogenous information. Furthermore, a siamese dual-branch backbone architecture captures dynamic patterns from the recomposed past and future representations, integrating them via a context-aware weighting mechanism to ensure dynamic balance. Extensive experiments on real-world datasets demonstrate the ExoST's effectiveness, universality, robustness, and efficiency. |
| title | Select, then Balance: Exploring Exogenous Variable Modeling of Spatio-Temporal Forecasting |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2509.05779 |