MoGERNN: An Inductive Traffic Predictor for Unobserved Locations in Dynamic Sensing Networks
This addresses a practical challenge in traffic management for cities with limited sensor coverage, offering an incremental improvement over existing deep learning methods by enabling predictions in unobserved areas.
The paper tackles the problem of predicting traffic states at unobserved locations in partially observed road networks, proposing MoGERNN, an inductive spatio-temporal graph representation model that outperforms baselines on real-world datasets and adapts to dynamic sensor configurations without costly retraining.
Given a partially observed road network, how can we predict the traffic state of unobserved locations? While deep learning approaches show exceptional performance in traffic prediction, most assume sensors at all locations of interest, which is impractical due to financial constraints. Furthermore, these methods typically require costly retraining when sensor configurations change. We propose MoGERNN, an inductive spatio-temporal graph representation model, to address these challenges. Inspired by the Mixture of Experts approach in Large Language Models, we introduce a Mixture of Graph Expert (MoGE) block to model complex spatial dependencies through multiple graph message aggregators and a sparse gating network. This block estimates initial states for unobserved locations, which are then processed by a GRU-based Encoder-Decoder that integrates a graph message aggregator to capture spatio-temporal dependencies and predict future states. Experiments on two real-world datasets show MoGERNN consistently outperforms baseline methods for both observed and unobserved locations. MoGERNN can accurately predict congestion evolution even in areas without sensors, offering valuable information for traffic management. Moreover, MoGERNN is adaptable to dynamic sensing networks, maintaining competitive performance even compared to its retrained counterpart. Tests with different numbers of available sensors confirm its consistent superiority, and ablation studies validate the effectiveness of its key modules.