Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting

TL;DR

The paper tackles spatiotemporal traffic forecasting by modeling spatial dependencies as diffusion processes on directed road graphs and integrating this with a sequence-to-sequence temporal framework. It introduces Diffusion Convolutional Recurrent Neural Network (DCRNN), which uses bidirectional diffusion convolutions (DCConv) within DCGRU units and a scheduled-sampling enhanced encoder-decoder to handle long-horizon predictions. Empirical results on METR-LA and PEMS-BAY show that DCRNN consistently outperforms a broad set of baselines, with larger gains for longer forecast horizons. The approach provides a principled, efficient way to capture directionality and non-Euclidean spatial structure in traffic networks and generalizes to other spatiotemporal forecasting tasks with evolving graphs.

Abstract

Spatiotemporal forecasting has various applications in neuroscience, climate and transportation domain. Traffic forecasting is one canonical example of such learning task. The task is challenging due to (1) complex spatial dependency on road networks, (2) non-linear temporal dynamics with changing road conditions and (3) inherent difficulty of long-term forecasting. To address these challenges, we propose to model the traffic flow as a diffusion process on a directed graph and introduce Diffusion Convolutional Recurrent Neural Network (DCRNN), a deep learning framework for traffic forecasting that incorporates both spatial and temporal dependency in the traffic flow. Specifically, DCRNN captures the spatial dependency using bidirectional random walks on the graph, and the temporal dependency using the encoder-decoder architecture with scheduled sampling. We evaluate the framework on two real-world large scale road network traffic datasets and observe consistent improvement of 12% - 15% over state-of-the-art baselines.

Figures (11)

• Figure 1: Spatial correlation is dominated by road network structure. (1) Traffic speed in road 1 are similar to road 2 as they locate in the same highway. (2) Road 1 and road 3 locate in the opposite directions of the highway. Though close to each other in the Euclidean space, their road network distance is large, and their traffic speeds differ significantly.
• Figure 2: System architecture for the Diffusion Convolutional Recurrent Neural Network designed for spatiotemporal traffic forecasting. The historical time series are fed into an encoder whose final states are used to initialize the decoder. The decoder makes predictions based on either previous ground truth or the model output.
• Figure 3: Learning curve for DCRNN and DCRNN without diffusion convolution. Removing diffusion convolution results in much higher validation error. Moreover, DCRNN with bi-directional random walk achieves the lowest validation error.
• Figure 4: Effects of K and the number of units in each layer of DCRNN. K corresponds to the reception field width of the filter, and the number of units corresponds to the number of filters.
• Figure 5: Performance comparison for different DCRNN variants. DCRNN, with the sequence to sequence framework and scheduled sampling, achieves the lowest MAE on the validation dataset. The advantage becomes more clear with the increase of the forecasting horizon.

Theorems & Definitions (2)

• Lemma 2.1
• Proposition 2.2