Graph Convolutional Networks for Traffic Forecasting with Missing Values

TL;DR

This paper introduces GCN-M, a novel graph convolutional network designed to accurately forecast traffic by effectively handling complex missing data scenarios in spatio-temporal traffic datasets.

Contribution

The paper presents a new GCN-based model that jointly addresses missing data imputation and traffic prediction using attention mechanisms and dynamic graph learning.

Findings

Outperforms existing models on real traffic datasets.

Effectively manages various missing data patterns.

Demonstrates robustness in complex missing-value scenarios.

Abstract

Traffic forecasting has attracted widespread attention recently. In reality, traffic data usually contains missing values due to sensor or communication errors. The Spatio-temporal feature in traffic data brings more challenges for processing such missing values, for which the classic techniques (e.g., data imputations) are limited: 1) in temporal axis, the values can be randomly or consecutively missing; 2) in spatial axis, the missing values can happen on one single sensor or on multiple sensors simultaneously. Recent models powered by Graph Neural Networks achieved satisfying performance on traffic forecasting tasks. However, few of them are applicable to such a complex missing-value context. To this end, we propose GCN-M, a Graph Convolutional Network model with the ability to handle the complex missing values in the Spatio-temporal context. Particularly, we jointly model the missing value processing and traffic forecasting tasks, considering both local Spatio-temporal features and global historical patterns in an attention-based memory network. We propose as well a dynamic graph learning module based on the learned local-global features. The experimental results on real-life datasets show the reliability of our proposed method.