TraverseNet: Unifying Space and Time in Message Passing for Traffic Forecasting

TL;DR

TraverseNet is a novel graph neural network that unifies space and time in traffic forecasting, capturing dynamic dependencies more effectively than previous models.

Contribution

It introduces a message traverse mechanism to model evolving spatial-temporal dependencies as an inseparable whole, improving traffic prediction accuracy.

Findings

TraverseNet outperforms existing models in traffic forecasting accuracy.

Ablation studies confirm the effectiveness of the message traverse mechanism.

Parameter studies demonstrate robustness across different settings.

Abstract

This paper aims to unify spatial dependency and temporal dependency in a non-Euclidean space while capturing the inner spatial-temporal dependencies for traffic data. For spatial-temporal attribute entities with topological structure, the space-time is consecutive and unified while each node's current status is influenced by its neighbors' past states over variant periods of each neighbor. Most spatial-temporal neural networks for traffic forecasting study spatial dependency and temporal correlation separately in processing, gravely impaired the spatial-temporal integrity, and ignore the fact that the neighbors' temporal dependency period for a node can be delayed and dynamic. To model this actual condition, we propose TraverseNet, a novel spatial-temporal graph neural network, viewing space and time as an inseparable whole, to mine spatial-temporal graphs while exploiting the evolving spatial-temporal dependencies for each node via message traverse mechanisms. Experiments with ablation and parameter studies have validated the effectiveness of the proposed TraverseNet, and the detailed implementation can be found from https://github.com/nnzhan/TraverseNet.