This book reports on important progress in the field of intelligent transportation systems. To solve non-trivial problems in the traffic state modeling and reliable route planning, this book proposes a series of novel methods: 1) a customized bidirectional recurrent neural network for non-original missing data imputation; 2) a spatiotemporal matrix factorization approach integrating traffic speed for original missing traffic flow data imputation; 3) a sequence-to-sequence learning model with graph convolution for multistep traffic state prediction; 4) a data organization scheme based on the road segment clustering to enhance traffic prediction efficiency; 5) a routing method to determine the least expected time paths considering uncertainty of traffic state prediction. These contents logically go forward one by one, which bridge the gap between raw detection data and smart mobility services. Finally, the author collects six large-scale real-world traffic datasets for numerical experiments. In contrast with over thirty benchmark models, the superiority of proposed approaches is validated.

Zhengchao Zhang is currently a Lecturer at the School of Computer Science and Technology, Soochow University, Suzhou, China. In 2022, he was awarded the outstanding doctoral degree by Tsinghua University, and also the best doctoral dissertation by Chinese Overseas Transportation Association. His research interests include intelligent transportation systems, smart mobility services, traffic data mining, autonomous driving, and deep learning.