{"slug": "emagn-efficient-multi-attention-graph-network-via-learned-clustering-for-traffic", "title": "EMAGN: Efficient Multi-Attention Graph Network via Learned Clustering for Scalable Traffic Forecasting", "summary": "Researchers propose EMAGN, an efficient multi-attention graph network that uses learned clustering to linearize spatial attention for traffic forecasting, reducing complexity from quadratic to linear. The model achieves accuracy within 2.7-3.2% MAE of full-attention GMAN while cutting training time by 32%, inference time by 38%, and GPU memory by 58%, and enables operation on standard GPUs where full-attention models fail.", "body_md": "arXiv:2607.13241v1 Announce Type: new\nAbstract: Traffic forecasting is highly challenging due to complex and nonlinear spatial and temporal dependencies. Self-attention mechanisms have been widely adopted to model dynamic and long-range dependencies, achieving state-of-the-art performance, but suffer from limited scalability due to quadratic computational and memory complexity. To address this, we propose an Efficient Multi-Attention Graph Network (EMAGN) that linearises the spatial attention mechanism itself, inspired by the theory of fast high-dimensional Gaussian filtering. Two learned clustering matrices C_k and C_v adaptively group key and value vectors into M super-clusters, reducing complexity from O(N^2 d) to O(NMd) without sacrificing the flexibility of attention for dynamic dependency modelling. Experimental results on PEMS-BAY and METR-LA show that EMAGN achieves accuracy within 2.7-3.2% MAE of full-attention GMAN while reducing training time by 32%, inference time by 38%, and GPU memory by 58%. Critically, at K=16 attention heads, full-attention GMAN runs out of memory on a standard 11 GB GPU entirely while EMAGN continues to operate, demonstrating a categorical expansion of feasible model configurations. EMAGN also surpasses Linformer and Performer in both accuracy and efficiency within the same backbone, owing to its traffic-network-aware adaptive clustering.", "url": "https://wpnews.pro/news/emagn-efficient-multi-attention-graph-network-via-learned-clustering-for-traffic", "canonical_source": "https://arxiv.org/abs/2607.13241", "published_at": "2026-07-16 04:00:00+00:00", "updated_at": "2026-07-16 04:29:38.753897+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "neural-networks", "ai-research"], "entities": ["EMAGN", "GMAN", "Linformer", "Performer", "PEMS-BAY", "METR-LA"], "alternates": {"html": "https://wpnews.pro/news/emagn-efficient-multi-attention-graph-network-via-learned-clustering-for-traffic", "markdown": "https://wpnews.pro/news/emagn-efficient-multi-attention-graph-network-via-learned-clustering-for-traffic.md", "text": "https://wpnews.pro/news/emagn-efficient-multi-attention-graph-network-via-learned-clustering-for-traffic.txt", "jsonld": "https://wpnews.pro/news/emagn-efficient-multi-attention-graph-network-via-learned-clustering-for-traffic.jsonld"}}