EMAGN: Efficient Multi-Attention Graph Network via Learned Clustering for Scalable Traffic Forecasting
For traffic forecasting practitioners, EMAGN offers a scalable attention mechanism that maintains accuracy while drastically reducing resource requirements, enabling larger model configurations.
EMAGN linearizes spatial attention in traffic forecasting via learned clustering, reducing complexity from O(N^2 d) to O(NMd). It achieves accuracy within 2.7-3.2% MAE of full-attention GMAN while cutting training time by 32%, inference by 38%, and GPU memory by 58%, and enables operation on 11 GB GPU where GMAN fails.
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.