Predicting Abnormal Events in Urban Rail Transit Systems with Multivariate Point Process

Mo, Xiaoyun; Li, Mingqian; Li, Zhenjiang

doi:10.1007/978-3-031-05933-9_4

Cited by 3 publications

(1 citation statement)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…first work tackles Challenge 1, non-recurrent mobility patterns, in traffic flow prediction, via spatiotemporal modeling with graph neural networks, which will be detailed in Chapter 3; our second work is motivated by Challenge 2, experimental limitations in actuation, in traffic safety education to reduce traffic accidents, to analyze the sample re-weighting techniques in uplift modeling, which will be detailed in Chapter 4. In addition, our collaborative work[80] tackles Challenge 1 in urban rail transit system via the modeling of multivariate point process; and our collaborative work[120] tackles Challenge 3 and 4 in reconstructing large-scale vehicle trajectories from camera sensing data via graph convolution, fusing vehicle identity information with vision-based information and spatiotemporal constraints by the road network. Details of these works can be found in our publications listed in Appendix A.Chapter 3 Traffic Flow Prediction with Vehicle Trajectories 1Our first work examines a problem in urban transport understanding, namely, vehicular traffic flow prediction.…”

mentioning

confidence: 99%

Machine learning models in urban transport applications: understanding and actuation

View full text Add to dashboard Cite

First and foremost, I would like to express my sincere gratitude to Professor Li Mo from the School of Computer Science and Engineering, for being my PhD supervisor in NTU and for his continuous support and guidance for my research projects. I would always appreciate his patience in guiding me through my PhD studies from brainstorming research topics to drafting papers, from rebuttal to presentation, from lab experiments to industrial collaborations, and from research to life, without which I would not make this thesis possible. I would also remember the times we partied at Prof. Li's house, with delicious food and fun activities. I am also grateful to my co-supervisor, Professor Cyril Leung, and my mentor, Dr. Wang Di, for being the members of my Thesis Advisory Committee and providing constructive feedback and insightful comments on my work via emails, talks and catch-ups. I would also like to thank my group mates in WANDS for their great support and encouragement. I would always remember the extensive discussion I had with Dr. Tong Panrong and Dr. Mo Xiaoyun on our collaborative work, the generous help from Dr. Cao Chu, Liu Kaiqi and Ji Sijie on every detail in research and in life. I would like to extend my thanks to my peers in Alibaba-NTU Talent Program for always being encouraging, positive and enthusiatic; and to the work mates in Alibaba for the valuable support and research facilities. Last but not least, I would like to thank my husband, Han Qiao, for being the greatest companion in my life, giving support for every decision I made and giving help for every struggle I had during my research and my life. I would like to thank my baby, Han Ziqian, for the happy time I had with him. I would also like to thank my parents, family, peers and friends for their love, support and encouragement. v Contents Statement of Originality i Supervisor Declaration Statement ii

show abstract

mentioning

confidence: 99%