Long-term Spatio-temporal Forecasting via Dynamic Multiple-Graph Attention

TL;DR

This paper introduces a novel dynamic multi-graph neural network model that effectively captures long-term spatio-temporal dependencies and contextual information, significantly enhancing prediction accuracy in long-term forecasting tasks.

Contribution

It proposes a new graph modeling and fusion approach with a dynamic multi-graph fusion module and trainable importance weights, addressing limitations of previous MGNN methods for LSTF.

Findings

Significant performance improvements over existing models on large-scale datasets.

Effective modeling of long-term dependencies and contextual information.

Enhanced prediction accuracy in long-term spatio-temporal forecasting.

Abstract

Many real-world ubiquitous applications, such as parking recommendations and air pollution monitoring, benefit significantly from accurate long-term spatio-temporal forecasting (LSTF). LSTF makes use of long-term dependency between spatial and temporal domains, contextual information, and inherent pattern in the data. Recent studies have revealed the potential of multi-graph neural networks (MGNNs) to improve prediction performance. However, existing MGNN methods cannot be directly applied to LSTF due to several issues: the low level of generality, insufficient use of contextual information, and the imbalanced graph fusion approach. To address these issues, we construct new graph models to represent the contextual information of each node and the long-term spatio-temporal data dependency structure. To fuse the information across multiple graphs, we propose a new dynamic multi-graph fusion module to characterize the correlations of nodes within a graph and the nodes across graphs via the spatial attention and graph attention mechanisms. Furthermore, we introduce a trainable weight tensor to indicate the importance of each node in different graphs. Extensive experiments on two large-scale datasets demonstrate that our proposed approaches significantly improve the performance of existing graph neural network models in LSTF prediction tasks.