A Convolution Component-Based Method with Attention Mechanism for Travel-Time Prediction.

Xiangdong Ran,Chuang Lin,Zhiguang Shan,Yufei Fang

doi:10.3390/s19092063

Abstract

Deep learning approaches have been recently applied to traffic prediction because of their ability to extract features of traffic data. While convolutional neural networks may improve the predictive accuracy by transiting traffic data to images and extracting features in the images, the convolutional results can be improved by using the global-level representation that is a direct way to extract features. The time intervals are not considered as aspects of convolutional neural networks for traffic prediction. The attention mechanism may adaptively select a sequence of regions and only process the selected regions to better extract features when aspects are considered. In this paper, we propose the attention mechanism over the convolutional result for traffic prediction. The proposed method is based on multiple links. The time interval is considered as the aspect of attention mechanism. Based on the dataset provided by Highways England, the experimental results show that the proposed method can achieve better accuracy than the baseline methods.

Highlights

In recent years, the increase in vehicle transit and congestion on highways and urban road networks has led the changes on traffic conditions to uncertainty
Deep learning approaches, such as the recurrent state–space neural network (SSNN) [4], the long short-term memory neural network (LSTM NN) [5], the autoencoders (AES) [6], the deep learning approach with a sequence of tanh layers [7] and CNN-based methods [8,9,10] have been proposed for traffic prediction in recent years with their abilities to extract effectively features from traffic data
Based on the dataset provided by Highways England, the proposed method is trained and the experimental results show that the proposed method can achieve better predictive accuracy over the baseline methods

Summary

Introduction

The increase in vehicle transit and congestion on highways and urban road networks has led the changes on traffic conditions to uncertainty. In various traffic indexes including traffic volume, travel-time, average speed, queue length and the severity of incidents, the delivery of travel-time is widely accepted as an index of ITS because travel-time is very intuitive and is understood [1,2,3] Deep learning approaches, such as the recurrent state–space neural network (SSNN) [4], the long short-term memory neural network (LSTM NN) [5], the autoencoders (AES) [6], the deep learning approach with a sequence of tanh layers [7] and CNN-based methods [8,9,10] have been proposed for traffic prediction in recent years with their abilities to extract effectively features from traffic data. CNN-based methods of traffic prediction may improve the predictive

Methods

Results

Discussion

Conclusion