Graph Convolutional Networks with Kalman Filtering for Traffic Prediction

Abstract

Traffic prediction is a challenging task due to the time-varying nature of traffic patterns and the complex spatial dependency of road networks. Adding to the challenge, there are a number of errors introduced in traffic sensor reporting, including bias and noise. However, most of the previous works treat the sensor observations as exact measures ignoring the effect of unknown noise. To model the spatial and temporal dependencies, existing studies combine graph neural networks (GNNs) with other deep learning techniques but their equal weighting of different dependencies limits the models' ability to capture the real dynamics in the traffic network. To deal with the above issues, we propose a novel deep learning framework called Deep Kalman Filtering Network (DKFN) to forecast the network-wide traffic state by modeling the self and neighbor dependencies as two streams, and their predictions are fused under the statistical theory and optimized through the Kalman filtering network. First, the reliability of each stream is evaluated using variances. Then, the Kalman filter is leveraged to properly fuse noisy observations in terms of their reliability. Experimental results reflect the superiority of the proposed method over baseline models on two real-world traffic datasets in the speed prediction task.

Graph Contrastive Learning with Cohesive Subgraph Awareness

“…Most existing GCL methods [17,58] use standard Graph Neural Networks (GNNs) such as GCN [22] and GIN [54] as GNN encoder. However, prior research has indicated that GNNs have limited expressive power and encounter difficulties in capturing subgraph properties [11]. Can we engineer a more expressive graph encoder that can effectively capture subgraph information from the original graph?…”

“…Although augmented graphs maintain a higher degree of cohesive subgraphs, the risk of losing subgraph information during the graph representation learning process remains. Current studies, such as [11], have underscored that plain GNNs struggle to accurately capture subgraph properties. To address this, inspired by [9], we then propose an original-graph-oriented graph substructure network (O-GSN) to enhance GNNs' power to aware graph cohesive substructures efficiently when encoding graphs.…”

A deep neural network model with GCN and 3D convolutional network for short‐term metro passenger flow forecasting

“…In recent years, some researchers have modelled the data on the nodes of a graph and forecasted based on the graph, using graph neural networks (GNNs) for forecasting. These models can capture topological information between stations to deal with graph-structured data for traffic forecasting [23,24]. DCRNN [25] regarded the traffic flow as a diffusion process on the graph and captures the spatial dependency with bidirectional random walks.…”

SUREL+: Moving from Walks to Sets for Scalable Subgraph-based Graph Representation Learning

“…Given a set of nodes of interest, namely a queried node-set, SGRL models such as SEAL [54,57], GraIL [41], and SubGNN [1] first extract a subgraph around the queried node-set (termed query-induced subgraph), and then encode the extracted subgraph for prediction. Extensive works have shown that SGRL models are more robust [52] and more expressive [5,12]; while canonical graph neural networks (GNNs) including GCN [23] and GraphSAGE [14] generally fail to make accurate predictions, due to their limited expressive power [9,13,57], incapability of capturing intra-node distance information [28,39], and improper entanglement between receptive field size and model depth [18,51,52]. An example in Fig.…”