A Lightweight and Accurate Spatial-Temporal Transformer for Traffic Forecasting
Guanyao Li, Shuhan Zhong, Xingdong Deng, Letian Xiang, S.-H. Gary Chan, Ruiyuan Li, Yang Liu, Ming Zhang, Chih‐Chieh Hung, Wen-Chih Peng
Abstract
We study the forecasting problem for traffic with dynamic, possibly periodical, and joint spatial-temporal dependency between regions. Given the aggregated inflow and outflow traffic of regions in a city from time slots 0 to <inline-formula><tex-math notation="LaTeX">$t - 1$</tex-math></inline-formula> , we predict the traffic at time <inline-formula><tex-math notation="LaTeX">$t$</tex-math></inline-formula> for any region. Prior arts in the area often considered the spatial and temporal dependencies in a decoupled manner, or were rather computationally intensive in training with a large number of hyper-parameters which needed tuning. We propose ST-TIS, a novel, lightweight and accurate <b>S</b> patial- <b>T</b> emporal <b>T</b> ransformer with <b>i</b> nformation fusion and region <b>s</b> ampling for traffic forecasting. ST-TIS extends the canonical Transformer with information fusion and region sampling. The information fusion module captures the complex spatial-temporal dependency between regions. The region sampling module is to improve the efficiency and prediction accuracy, cutting the computation complexity for dependency learning from <inline-formula><tex-math notation="LaTeX">$O(n^{2})$</tex-math></inline-formula> to <inline-formula><tex-math notation="LaTeX">$O(n\sqrt{n})$</tex-math></inline-formula> , where <inline-formula><tex-math notation="LaTeX">$n$</tex-math></inline-formula> is the number of regions. With far fewer parameters than state-of-the-art deep learning models, ST-TIS's offline training is significantly faster in terms of tuning and computation (with a reduction of up to <inline-formula><tex-math notation="LaTeX">$90\%$</tex-math></inline-formula> on training time and network parameters). Notwithstanding such training efficiency, extensive experiments show that ST-TIS is substantially more accurate in online prediction than state-of-the-art approaches (with an average improvement of <inline-formula><tex-math notation="LaTeX">$9.5\%$</tex-math></inline-formula> on RMSE, and <inline-formula><tex-math notation="LaTeX">$12.4\%$</tex-math></inline-formula> on MAPE compared to STDN and DSAN).