aiqbench analysis
AdaST introduces a novel framework for spatial-temporal forecasting by dynamically adjusting model components based on whether data patterns are primarily spatial, temporal, or strongly coupled. This adaptive approach addresses the limitations of existing methods that assume uniform correlation structures, leading to improved accuracy in complex real-world systems like traffic and climate networks. By explicitly handling heterogeneous coupling dynamics, the research advances the state-of-the-art in machine learning for time-series prediction tasks.
AdaST: Adaptive Coupling for Spatial-Temporal Forecasting
arXiv:2609.36119v1 Announce Type: new Abstract: Spatial-temporal (ST) forecasting underpins many real-world systems such as traffic, climate, and energy networks. While existing methods implicitly assume strong spatiotemporal coupling, we observe that real-world ST data exhibits distinct coupling regimes, ranging from temporal-dominated and spatial-dominated to strongly coupled patterns. This mismatch causes current models to suffer from spurious dependencies and degraded performance when one correlation dominates. To overcome this limitation, we aim to dynamically modulate spatial and temporal modeling based on the data's inherent coupling structure. However, three key challenges exist: unknown coupling structure, heterogeneous coupling dynamics, and suboptimal spatial modeling. We propose AdaST, an adaptive ST forecasting framework that tackles these challenges through a decompose-recompose paradigm. AdaST factorizes inputs into components capturing different coupling patterns using heterogeneity-aware experts. Each component is processed by role-aligned modules, and a correlation-informed adaptive recomposer integrates them for final prediction. Extensive experiments confirm that AdaST significantly outperforms state-of-the-art baselines, validating the necessity of an adaptive approach.
Abstract:Spatial-temporal (ST) forecasting underpins many real-world systems such as traffic, climate, and energy networks. While existing methods implicitly assume strong spatiotemporal coupling, we observe that real-world ST data exhibits distinct coupling regimes, ranging from temporal-dominated and spatial-dominated to strongly coupled patterns. This mismatch causes current models to suffer from spurious dependencies and degraded performance when one correlation dominates. To overcome this limitation, we aim to dynamically modulate spatial and temporal modeling based on the data's inherent coupling structure. However, three key challenges exist: unknown coupling structure, heterogeneous coupling dynamics, and suboptimal spatial modeling. We propose AdaST, an adaptive ST forecasting framework that tackles these challenges through a decompose-recompose paradigm. AdaST factorizes inputs into components capturing different coupling patterns using heterogeneity-aware experts. Each component is processed by role-aligned modules, and a correlation-informed adaptive recomposer integrates them for final prediction. Extensive experiments confirm that AdaST significantly outperforms state-of-the-art baselines, validating the necessity of an adaptive approach.
Accepted to NeurIPS 2026 (main conference). Code: this https URL
Artificial Intelligence (cs.AI)
arXiv:2609.36119 [cs.AI]
(or arXiv:2609.36119v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2609.36119 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 258 words
Related coverage
DSFNet: Learning Dual-Domain Spectral Operators for Multi-Modality Spatio-Temporal Forecasting in Urban Transportation Systems
arXiv · Jun 9, 2026
VaryST: Generalizable Spatio-Temporal Forecasting Beyond the Closed-World Setting
arXiv · Oct 15, 2026
LLM-enhanced spatio-temporal learning for grid-level docked bike sharing demand prediction
arXiv · Oct 6, 2026
UniST-Pred: A Robust Unified Framework for Spatio-Temporal Traffic Forecasting in Transportation Networks Under Disruptions
arXiv · Oct 1, 2026