{"slug": "structure-guided-spatiotemporal-attention-graph-neural-network-for-traffic-flow", "title": "Structure-Guided Spatiotemporal Attention Graph Neural Network for Traffic Flow Prediction", "summary": "Researchers propose the Structure-Guided Spatiotemporal Attention Graph Neural Network (SGSAN) for traffic flow prediction, achieving state-of-the-art accuracy while providing built-in interpretability. The model learns a static Directed Dependency Graph (DDG) and uses an InfoNCE-based soft-coupling mechanism to align attention with macroscopic traffic dependencies, addressing the lack of transparency in existing deep spatiotemporal models. Experiments on real-world datasets show SGSAN outperforms current methods in predictive accuracy and interpretability.", "body_md": "arXiv:2608.14177v1 Announce Type: new\nAbstract: Deep spatiotemporal models integrating graph convolutions and attention mechanisms have demonstrated excellent performance in network-level traffic flow prediction, owing to their exceptional ability to capture complex spatiotemporal dependencies. Despite their predictive success, deployment of such models in safety-critical urban systems remains constrained by their inherent lack of transparency. Existing post-hoc diagnostic methods often struggle with spurious correlations and fail to unveil the intrinsic decision-making mechanisms governing traffic dynamics, resulting in suboptimal interpretability and limited operational trustworthiness. To address these challenges, this paper proposes the Structure-Guided Spatiotemporal Attention Graph Neural Network (SGSAN). Departing from traditional architectures that rely on unconstrained adaptive graphs, SGSAN explicitly learns a static Directed Dependency Graph (DDG) to identify the invariant macroscopic propagation paths of traffic states. We further introduce an InfoNCE-based soft-coupling mechanism that anchors the model's dynamic spatiotemporal attention to this structural prior, offering a mechanistic account of the model's decision-making process while ensuring robust forecasting by aligning attention-based reasoning with identified macroscopic dependencies and preventing over-reliance on ephemeral local noise. Furthermore, a decoupled two-stage optimization framework is developed to resolve the fundamental conflict between structural discovery and predictive error minimization. Extensive experiments on multiple real-world datasets demonstrate that SGSAN achieves state-of-the-art predictive accuracy while providing built-in interpretability that organically aligns with the physical logic of traffic networks.", "url": "https://wpnews.pro/news/structure-guided-spatiotemporal-attention-graph-neural-network-for-traffic-flow", "canonical_source": "https://www.machinebrief.com/news/structure-guided-spatiotemporal-attention-graph-neural-netwo-hsfv", "published_at": "2026-08-17 04:00:00+00:00", "updated_at": "2026-08-17 04:42:24.316240+00:00", "lang": "en", "topics": ["machine-learning", "artificial-intelligence"], "entities": ["SGSAN", "InfoNCE"], "alternates": {"html": "https://wpnews.pro/news/structure-guided-spatiotemporal-attention-graph-neural-network-for-traffic-flow", "markdown": "https://wpnews.pro/news/structure-guided-spatiotemporal-attention-graph-neural-network-for-traffic-flow.md", "text": "https://wpnews.pro/news/structure-guided-spatiotemporal-attention-graph-neural-network-for-traffic-flow.txt", "jsonld": "https://wpnews.pro/news/structure-guided-spatiotemporal-attention-graph-neural-network-for-traffic-flow.jsonld"}}