Capturing Complex Spatial-Temporal Dependencies in Traffic Forecasting: A Self-Attention Approach

TL;DR

This paper introduces ST-SAM, a self-attention based model that effectively captures complex joint spatial-temporal dependencies in traffic forecasting, outperforming existing methods in accuracy and efficiency.

Contribution

The paper proposes a novel self-attention based model, ST-SAM, that jointly captures spatial and temporal dependencies for traffic forecasting, improving accuracy and efficiency over prior decoupled approaches.

Findings

ST-SAM outperforms state-of-the-art methods with up to 15% lower RMSE.

ST-SAM achieves 17% better MAPE.

Training time is reduced by 32 times compared to previous models.

Abstract

We study the problem of traffic forecasting, aiming to predict the inflow and outflow of a region in the subsequent time slot. The problem is complex due to the intricate spatial and temporal interdependence among regions. Prior works study the spatial and temporal dependency in a decouple manner, failing to capture their joint effect. In this work, we propose ST-SAM, a novel and efficient Spatial-Temporal Self-Attention Model for traffic forecasting. ST-SAM uses a region embedding layer to learn time-specific embedding from traffic data for regions. Then, it employs a spatial-temporal dependency learning module based on self-attention mechanism to capture the joint spatial-temporal dependency for both nearby and faraway regions. ST-SAM entirely relies on self-attention to capture both local and global spatial-temporal correlations, which make it effective and efficient. Extensive experiments on two real world datasets show that ST-SAM is substantially more accurate and efficient than the state-of-the-art approaches (with an average improvement of up to 15% on RMSE, 17% on MAPE, and 32 times on training time in our experiments).