arXiv:2505.17637v2 Announce Type: replace 
Abstract: Spatio-temporal prediction plays a crucial role in intelligent transportation, weather forecasting, and urban planning. While integrating multi-modal data has shown potential for enhancing prediction accuracy, key challenges persist: (i) inadequate fusion of multi-modal information, (ii) confounding factors that obscure causal relations, and (iii) high computational complexity of prediction models. To address these challenges, we propose E^2-CSTP, an Effective and Efficient Causal multi-modal Spatio-Temporal Prediction framework. E^2-CSTP leverages cross-modal attention and gating mechanisms to effectively integrate multi-modal data. Building on this, we design a dual-branch causal inference approach: the primary branch focuses on spatio-temporal prediction, while the auxiliary branch mitigates bias by modeling additional modalities and applying causal interventions to uncover true causal dependencies. To improve model efficiency, we integrate GCN with the Mamba architecture for accelerated spatio-temporal encoding. Extensive experiments on 4 real-world datasets show that E^2-CSTP significantly outperforms 9 state-of-the-art methods, achieving up to 9.66% improvements in accuracy as well as 17.37%-56.11% reductions in computational overhead.

تم اقتراح نهج جديد يسمى E^2-C لتحسين التنبؤ الزماني المكاني، وهو أمر حيوي في مجالات مثل النقل الذكي وتوقعات الطقس. يتناول هذا الأسلوب تحديات كبيرة مثل دمج البيانات متعددة الأنماط وتعقيد نماذج التنبؤ. من خلال تحسين دقة التنبؤات، يمكن أن يؤدي E^2-C إلى اتخاذ قرارات أفضل في التخطيط الحضري ومجالات حيوية أخرى، مما يجعله تقدمًا ملحوظًا في هذا المجال.

Se ha propuesto un nuevo enfoque llamado E^2-C para mejorar la predicción espaciotemporal, que es vital en campos como el transporte inteligente y la previsión meteorológica. Este método aborda desafíos significativos como la fusión de datos multimodales y la complejidad de los modelos de predicción. Al mejorar la precisión de las predicciones, E^2-C podría llevar a una mejor toma de decisiones en la planificación urbana y otras áreas críticas, lo que lo convierte en un avance notable en el campo.

Une nouvelle approche appelée E^2-C a été proposée pour améliorer la prédiction spatio-temporelle, qui est essentielle dans des domaines comme le transport intelligent et la prévision météorologique. Cette méthode répond à des défis majeurs tels que la fusion de données multimodales et la complexité des modèles de prédiction. En améliorant la précision des prévisions, E^2-C pourrait conduire à de meilleures prises de décision dans la planification urbaine et d'autres domaines critiques, ce qui en fait une avancée remarquable dans le domaine.

A new approach called E^2-C has been proposed to improve spatio-temporal prediction, which is vital for fields like intelligent transportation and weather forecasting. This method addresses significant challenges such as the fusion of multi-modal data and the complexity of prediction models. By enhancing prediction accuracy, E^2-C could lead to better decision-making in urban planning and other critical areas, making it a noteworthy advancement in the field.

Causal Spatio-Temporal Prediction: An Effective and Efficient Multi-Modal Approach

One More Thing in AI – Your Shortcut to AI Mastery

Causal Spatio-Temporal Prediction: An Effective and Efficient Multi-Modal Approach

Was this article worth reading? Share it

One More Thing in AI

LucidQuery AI

Attentive AI

Https

Meteoria

Spentrip

Ready to build your own newsroom?