Abstract:Research in deep learning models to forecast traffic intensities has gained great attention in recent years due to their capability to capture the complex spatio-temporal relationships within the traffic data. However, most state-of-the-art approaches have designed spatial-only (e.g. Graph Neural Networks) and temporal-only (e.g. Recurrent Neural Networks) modules to separately extract spatial and temporal features. However, we argue that it is less effective to extract the complex spatio-temporal relationship with such factorized modules. Besides, most existing works predict the traffic intensity of a particular time interval only based on the traffic data of the previous one hour of that day. And thereby ignores the repetitive daily/weekly pattern that may exist in the last hour of data. Therefore, we propose a Unified Spatio-Temporal Graph Convolution Network (USTGCN) for traffic forecasting that performs both spatial and temporal aggregation through direct information propagation across different timestamp nodes with the help of spectral graph convolution on a spatio-temporal graph. Furthermore, it captures historical daily patterns in previous days and current-day patterns in current-day traffic data. Finally, we validate our work's effectiveness through experimental analysis, which shows that our model USTGCN can outperform state-of-the-art performances in three popular benchmark datasets from the Performance Measurement System (PeMS). Moreover, the training time is reduced significantly with our proposed USTGCN model.

What problem does this paper attempt to address?

This paper attempts to solve the problem of modeling complex spatio - temporal relationships in traffic flow prediction. Specifically, the paper points out that most of the current state - of - the - art methods use spatial modules (such as graph neural networks) and temporal modules (such as recurrent neural networks) to extract the spatial and temporal features of traffic data respectively, and this separated method is not effective in capturing complex spatio - temporal relationships. In addition, most of the existing work only predicts the traffic intensity within a specific time interval based on the data of the previous hour, ignoring the repeated daily/weekly patterns that may exist in the data of the past few hours. To solve these problems, the paper proposes a unified spatio - temporal graph convolutional network (USTGCN) for traffic prediction. USTGCN simultaneously performs spatial and temporal aggregations by directly spreading information between nodes at different timestamps, thereby more comprehensively learning the mutual relationships between neighboring nodes at different times. In addition, USTGCN also captures the historical daily patterns of the previous days and the current pattern of the current day to improve prediction accuracy. The main contributions of the paper are as follows: 1. Propose a new unified spatio - temporal graph convolutional network (USTGCN) that can capture the complex cross - spatio - temporal dependency relationships in traffic network data. 2. Propose a simple but effective method to extract the current pattern and historical daily patterns by analyzing the traffic data of the current day and the past week. 3. Experimental analysis shows that the proposed unified spatio - temporal model achieves state - of - the - art performance on three public datasets from the Performance Measurement System (PeMS). Through these innovations, USTGCN not only improves the accuracy of traffic prediction, but also significantly reduces the training time.

Unified Spatio-Temporal Modeling for Traffic Forecasting using Graph Neural Network

SST-GNN: Simplified Spatio-temporal Traffic forecasting model using Graph Neural Network

Probabilistic spatio-temporal graph convolutional network for traffic forecasting

Attention-Based Multiple Graph Convolutional Recurrent Network for Traffic Forecasting

A Graph Convolutional Stacked Temporal Attention Neural Network for Traffic Flow Forecasting

Attention Based Spatial-Temporal Graph Convolutional Networks for Traffic Flow Forecasting

Attention-based Spatial-Temporal Graph Convolutional Recurrent Networks for Traffic Forecasting

Graph learning-based spatial-temporal graph convolutional neural networks for traffic forecasting

Short-term Traffic Speed Forecasting Based on Graph Attention Temporal Convolutional Networks

DMGCRN: Dynamic Multi-Graph Convolution Recurrent Network for Traffic Forecasting

Spatial-Temporal Residual Multi-Graph Convolution Network for Traffic Forecasting

Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting

Spatio-Temporal Joint Graph Convolutional Networks for Traffic Forecasting

Graph Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting

Reinforced Spatiotemporal Attentive Graph Neural Networks for Traffic Forecasting.

Temporal Graph Convolutional Network for Urban Traffic Flow Prediction Method.

GSTC-Unet: A U-shaped multi-scaled spatiotemporal graph convolutional network with channel self-attention mechanism for traffic flow forecasting

Spatial-Temporal Hierarchical Graph Convolutional Networks for Traffic Forecasting

Traffic Graph Convolutional Recurrent Neural Network: A Deep Learning Framework for Network-Scale Traffic Learning and Forecasting

Bayesian Spatio-Temporal Graph Convolutional Network for Traffic Forecasting

Attention Mechanism Based on Temporal Graph Convolutional Neural Network for Traffic Flow Prediction