MSGAT-GRU: A Multi-Scale Graph Attention and Recurrent Model for Spatiotemporal Road Accident Prediction
This addresses road safety for urban traffic management, offering a scalable and interpretable model, though it appears incremental as it builds on existing graph attention and recurrent architectures.
The paper tackles road accident prediction by proposing MSGAT-GRU, a model that captures multi-scale spatial dependencies and temporal dynamics using heterogeneous inputs like traffic flow and weather, achieving an RMSE of 0.334 and F1-score of 0.878 on the Beijing dataset and demonstrating transferability with RMSE of 6.48 on METR-LA.
Accurate prediction of road accidents remains challenging due to intertwined spatial, temporal, and contextual factors in urban traffic. We propose MSGAT-GRU, a multi-scale graph attention and recurrent model that jointly captures localized and long-range spatial dependencies while modeling sequential dynamics. Heterogeneous inputs, such as traffic flow, road attributes, weather, and points of interest, are systematically fused to enhance robustness and interpretability. On the Hybrid Beijing Accidents dataset, MSGAT-GRU achieves an RMSE of 0.334 and an F1-score of 0.878, consistently outperforming strong baselines. Cross-dataset evaluation on METR-LA under a 1-hour horizon further supports transferability, with RMSE of 6.48 (vs. 7.21 for the GMAN model) and comparable MAPE. Ablations indicate that three-hop spatial aggregation and a two-layer GRU offer the best accuracy-stability trade-off. These results position MSGAT-GRU as a scalable and generalizable model for intelligent transportation systems, providing interpretable signals that can inform proactive traffic management and road safety analytics.