Multicomponent Spatial-Temporal Graph Attention Convolution Networks for Traffic Prediction with Spatially Sparse Data.

Multicomponent Spatial-Temporal Graph Attention Convolution Networks for Traffic Prediction with Spatially Sparse Data.
复制标题

用于空间稀疏数据流量预测的多分量时空图注意力卷积网络

DOI:
10.1155/2021/9134942
复制
发表时间:
2021
影响因子:
--
通讯作者:
Zhang H
Zhang H
中科院分区:
工程技术3区
文献类型:
--
作者:
Liu S;Dai S;Sun J;Mao T;Zhao J;Zhang H

文献摘要

参考文献

相似文献

预测流量网络上的流量数据对于运输管理至关重要。由于复杂的时空依赖性,这是一项具有挑战性的任务。最新研究主要集中于使用空间密集的交通数据捕获时间和空间依赖性。但是,当流量数据在空间上稀疏时,现有方法无法捕获足够的空间相关信息,因此无法充分学习时间周期性。为了解决这些问题,我们提出了一个新颖的深度学习框架,即多组分时空图形注意力卷积网络(MSTGACN)进行交通预测,我们成功地将其应用于通过空间稀疏数据来预测交通流和速度。 MSTGACN主要由三个独立组件组成,用于建模三种类型的周期性信息。 MSTGACN中的每个组件都结合了扩张的因果卷积,图形卷积层和重量共享的图表层。在空间稀疏数据的情况下,对三个现实世界流量数据集(Metr-LA,PEMS-Bay和PEMSD7-SPARSE)的实验结果证明了我们方法的出色性能。
Predicting traffic data on traffic networks is essential to transportation management. It is a challenging task due to the complicated spatial-temporal dependency. The latest studies mainly focus on capturing temporal and spatial dependencies with spatially dense traffic data. However, when traffic data become spatially sparse, existing methods cannot capture sufficient spatial correlation information and thus fail to learn the temporal periodicity sufficiently. To address these issues, we propose a novel deep learning framework, Multi-component Spatial-Temporal Graph Attention Convolutional Networks (MSTGACN), for traffic prediction, and we successfully apply it to predicting traffic flow and speed with spatially sparse data. MSTGACN mainly consists of three independent components to model three types of periodic information. Each component in MSTGACN combines dilated causal convolution, graph convolution layer, and the weight-shared graph attention layer. Experimental results on three real-world traffic datasets, METR-LA, PeMS-BAY, and PeMSD7-sparse, demonstrate the superior performance of our method in the case of spatially sparse data.
DOI: 10.1016/j.neucom.2020.06.001
发表时间: 2020-10-14
期刊: NEUROCOMPUTING
影响因子: 6
作者:
Guo, Ge;Yuan, Wei
通讯作者: Yuan, Wei
DOI: 10.1080/10826068.2019.1658119
发表时间: 2019-08-26
影响因子: 2.9
作者:
Wu, Zhaoliang;Yin, Hao;Zhao, Xiaomei
通讯作者: Zhao, Xiaomei
基于 LSTM 的缺失数据交通流预测
DOI: 10.1016/j.neucom.2018.08.067
发表时间: 2018-11-27
期刊: NEUROCOMPUTING
影响因子: 6
作者:
Tian, Yan;Zhang, Kaili;Yang, Bailin
通讯作者: Yang, Bailin
短期交通流量预测: 时间序列分析与监督学习的实验比较
DOI: 10.1109/tits.2013.2247040
发表时间: 2013-06-01
影响因子: 8.5
作者:
Lippi, Marco;Bertini, Matteo;Frasconi, Paolo
通讯作者: Frasconi, Paolo
DOI: 10.1109/tits.2004.837813
发表时间: 2004-12-01
影响因子: 8.5
作者:
Wu, CH;Ho, JM;Lee, DT
通讯作者: Lee, DT