The Development of Traffic Flow Prediction Based on Deep Learning: A Literature Review

Yongjia Lei, Zhiying Cui, Linyan Dai, Ning Xiao, Xiangjie Wang, Xiangfang Ma, Ying Hong

Abstract: As is well acknowledged, deep learning has been fairly widely used in both academic and industrial fields in recent periods. Researchers favour its strong portability and good performance. Although deep learning has shown satisfactory performance in traffic flow prediction, its specific development process in such a field is not very clear, which is also not conducive for researchers to choose and adjust the model. This paper mainly retraces some significant models applied in traffic prediction before and after the advent of neural networks and digs out the reasons for their appearance. Through review and mining, we have sorted out the development history of traffic flow prediction models, providing us with some inspiration for subsequent research.