1pm-5pm, August 5, 2019 in Anchorage, Alaska USA

2nd Workshop on Data-driven Intelligent Transportation (DIT 2019)

Held in conjunction with KDD 2019

Location: Summit 4 - Ground level, Egan Center

Transportation is one of the most important components in any city in the world. In US, transportation sector accounts for 29% of energy consumption and traffic congestion cost is $305 billion in 2017. In China, Tier 1, Tier 2, and Tier 3 cities are now “full of cars.” In 2017, Beijing had car ownership penetration of 260 cars per 1,000 people, while the top 10 cities averaged 251 cars per 1,000 people. An intelligent transportation system is much needed to enable sustainable and efficient city.

In the meantime, modern technologies enable us to collect city data at an unprecedented speed. A wide range of city data has become increasingly available, such as taxi trips, surveillance camera data, human mobility data from mobile phones or location-based services, events, car accidents, shared bikes, POI, traffic from loop sensors, public transportation data, and many more.

How can we utilize such large-scale city data towards a more intelligent transportation system? While intelligent transportation is not a new topic, especially in the field of transportation research, existing transportation is not effectively utilizing large-scale city data and new computing technologies. Data and computing can help tackle many transportation questions including traffic signal control, route planning, shared transportation, autonomous driving, mixed transportation environment and data sensing.

This workshop would like to bring together the researchers to share the exciting data-driven techniques to solve transportation problems.

Topics of interest include but not limited to:

  •     - Reinforcement learning for transportation
  •     - Traffic forecasting
  •     - Route planning
  •     - Travel time estimation
  •     - Traffic signal control
  •     - Shared transportation
  •     - Autonomous driving vehicles
  •     - City-wide traffic estimation
  •     - Semantic mobility data understanding
  •     - Large-scale city data analysis and modeling
  •     - Large-scale traffic data visualization and interactive design
  •     - Sustainable transportation system
  •     - City data sensing and collecting
  •     - City data fusion and mining
  •     - Anomaly detection and forecasting

This year, we will have 4 invited talks and 11 lightning talks and poster presentations. We will also assemble a panel to award the best poster in our workshop. Stay tuned for more details.

WORKSHOP ORGANIZERS

Image

Zhenhui (Jessie) Li Pennsylvania State University

Image

Sanjay Chawla Qatar Computing Research Institute

Image

Dimitrios Gunopulos University of Athens

Image

Cathy Wu MIT

Agenda

Opening Remarks

Invited Talk 1: Robust Machine Learning for Tomorrow’s Transportation Needs

Dragos D. Margineantu Boeing Research & Technology

Lightning Talks of Posters

Coffee Break and Poster Session

Invited Talk 2: Deep Reinforcement Learning in Ride-sharing Marketplace

Zhiwei (Tony) Qin Didi Chuxing

Invited Talk 3: PittSmartLiving: Rethinking Public Transportation through Incentives

Konstantinos Pelechrinis University of Pittsburgh

Invited Talk 4: Reinforcement Learning for Intelligent Transportation

Zhenhui (Jessie) Li Pennsylvania State University

Panel and Best Poster Award

Invited talks

Image

Dr. Dragos Margineantu

Boeing Research & Technology

Robust Machine Learning for Tomorrow’s Transportation Needs

Speaker: Dragos Margineantu

TBD

Bio: Dragos Margineantu is the AI Chief Technologist and a Technical Fellow of Boeing Research & Technology. His research interests include robust machine learning, anomaly detection, inverse reinforcement learning, decision systems, human-in-the-loop learning, validation and testing of decision systems, cost-sensitive, active, and ensemble learning.
Dragos was one of the research pioneers in ensemble learning and cost-sensitive learning. At Boeing, he designed and developed machine learning and AI based solutions for airplane maintenance, autonomous systems, surveillance, and design. Dragos is the Boeing AI lead for the DARPA “Assured Autonomy” program, focusing on robust machine learning techniques for autonomous systems. He also served as PI of DARPA's "Learning Applied to Ground Robots" and “Bootstrapped Learning” programs.
Dragos serves as the Editor of the Springer book series on “Applied Machine Learning” and as the Action Editor for Special Issues for the Machine Learning Journal (MLj). He serves on the editorial board of both major machine learning journals (MLj and JMLR), and served as senior program committee member of all major machine learning and AI research conferences. He was the chair of the KDD 2015 Industry and Government Track.
Image

Dr. Zhiwei (Tony) Qin

Didi Chuxing

Deep Reinforcement Learning in Ride-sharing Marketplace

Speaker: Zhiwei (Tony) Qin

With the rising prevalence of smart mobile phones in our daily life, online ride-hailing platforms have emerged as a viable solution to provide more timely and personalized transportation service, led by such companies as DiDi, Uber, and Lyft. These platforms also allow idle vehicle vacancy to be more effectively utilized to meet the growing need of on-demand transportation, by connecting potential mobility requests to eligible drivers. In this talk, we will discuss our train of research on ride-hailing marketplace optimization at DiDi, in particular, order dispatching and driver repositioning. We will show single-agent and multi-agent RL formulations and how value functions can be designed to leverage different amount of information.

Bio: Dr. Zhiwei (Tony) Qin leads the reinforcement learning research at DiDi AI Labs, working on core problems in ride-sharing marketplace optimization. He received his Ph.D. in Operations Research from Columbia University and B.Sc. in Computer Science and Statistics from the University of British Columbia, Vancouver. Tony is broadly interested in research topics at the intersection of optimization and machine learning, and most recently in reinforcement learning and its applications in operational optimization, digital marketing, traffic signals control, and education. He has published in top-tier conferences and journals in machine learning and optimization, including ICML, KDD, IEEE ICDM, WWW, JMLR, and MPC. He has served as Senior PC/PC of NeurIPS, AAAI, IJCAI, KDD, JMLR, TPAMI, and select operations research journals.
Image

Dr. Konstantinos Pelechrinis

University of Pittsburgh

PittSmartLiving: Rethinking Public Transportation through Incentives

Speaker: Konstantinos Pelechrinis

"An advanced nation is one where the rich people use public transit, and not one, where the poor have cars" said Enrique Penalosa, the celebrated mayor of Bogota. Inspired by this idea, in this project we rethink the ways to improve the public transit experience and convert more auto commuters to public transit. We a holistic view of all the stakeholders involved, namely, transit operator, commuter and local businesses (the latter rely on transportation systems to bring in customers) and we look among other things into what types of information will help commuters make the transition to public transit, how local businesses can help improve the public transit experience and how we can make this without increasing inequalities in the population.

Bio: Kostas is an associate professor at the School of Computing and Information at the University of Pittsburgh. He received his diploma from the Department of Electrical and Computer Engineering of the National Technical University of Athens, while he holds a PhD degree from the Department of Computer Science of the University of California, at Riverside. His research interests include applied machine learning and data science, with an emphasis on applications in urban science and sports. He has received the prestigious Young Investigator Award from the Army Research Office for his research on multidimensional networks, while he has also consulted for professional sports teams.
Image

Dr. Zhenhui (Jessie) Li

Pennsylvania State University

Reinforcement Learning for Intelligent Transportation

Speaker: Zhenhui (Jessie) Li

Large-scale mobility data can be collected from mobile phones, car navigation systems, road surveillance cameras, and loop sensors. Turning such mobility data into knowledge can provide insights about our city and empower the city to be more intelligent. This talk presents how to utilize mobility data and advanced learning methods for traffic signal control. First, we examine the existing traffic signal control system and discuss why today we have the opportunity for a potential breakthrough in traffic signal control. Second, the talk presents our recent research results in traffic signal control via deep reinforcement learning. We demonstrate how the classical transportation methods can be integrated to guide our reinforcement learning approach. Finally, we would like to discuss the open challenges in this research topic and share the experience in the field experiments of controlling traffic signals in the city of Hangzhou.

Dr. Zhenhui (Jessie) Li is a tenured associate professor of Information Sciences and Technology at the Pennsylvania State University. She is Haile family early career endowed professor. Prior to joining Penn State, she received her PhD degree in Computer Science from University of Illinois Urbana-Champaign in 2012, where she was a member of data mining research group. Her research has been focused on mining spatial-temporal data with applications in transportation, ecology, environment, social science, and urban computing. She is a passionate interdisciplinary researcher and has been actively collaborating with cross-domain researchers. She has served as organizing committee or senior program committee of many conferences including KDD, ICDM, SDM, CIKM, and SIGSPATIAL. She has received NSF CAREER award, junior faculty excellence in research, and George J. McMurtry junior faculty excellence in teaching and learning award. To learn more, please visit her homepage: https://faculty.ist.psu.edu/jessieli

Lightning Talks and Posters

- Short and Long-term Pattern Discovery Over Large-Scale Geo-Spatiotemporal Data

1:40 PM - 1:45PM

Sobhan Moosavi (The Ohio State University); Mohammad Hossein Samavatian (The Ohio State University); Arnab Nandi (The Ohio State University); Srinivasan Parthasarathy (The Ohio State University); Rajiv Ramnath (The Ohio State University)

- Origin-Destination Matrix Prediction via Graph Convolution: A New Perspective of Passenger Demand Modeling

1:45 PM - 1:50PM

Yuandong Wang (Beihang University);Hongzhi Yin (The University of Queensland);Hongxu Chen (The University of Queensland);Tianyu Wo (Beihang University);Jie Xu (University of Leeds);Kai Zheng (University of Electronic Science and Technology);

- A Deep Value-network Based Approach for Multi-Driver Order Dispatching

1:50 PM - 1:55PM

Xiaocheng Tang (Didi Chuxing), Zhiwei Qin (Didi Chuxing), Fan Zhang, Zhaodong Wang, Zhe Xu (Didi Chuxing), Yintai Ma, Hongtu Zhu and Jieping Ye (Didi Chuxing)

- Co-Prediction of Multiple Transportation Demands Based on Deep Spatio-Temporal Neural Network

1:55 PM - 2:00PM

Junchen Ye (SKLSDE Lab and BDBC Beihang University);Leilei Sun (SKLSDE Lab and BDBC Beihang University);Bowen Du (SKLSDE Lab and BDBC Beihang University);Yanjie Fu (Missouri University of Science and Technology);Xinran Tong (SKLSDE Lab and BDBC Beihang University);Hui Xiong (Rutgers Business School Rutgers University);

- Empowering A* Search Algorithms with Neural Networks for Personalized Route Recommendation

2:00 PM - 2:05PM

Jingyuan Wang (Beihang University);Ning Wu (Beihang University);Xin Zhao (Renmin University of China);Fanzhang Peng (Beihang University);Xin Lin (Beihang University);

- Time Critic Policy Gradient Methods for Traffic Signal Control in Complex and Congested Scenarios

2:05 PM - 2:10PM

Stefano Giovanni Rizzo (Qatar Computing Research Institute);Giovanna Vantini (Qatar Computing Research Institute);Sanjay Chawla (Qatar Computing Research Institute);

- PressLight: Learning Max Pressure Control to Coordinate Traffic Signals in Arterial Network

2:10 PM - 2:15PM

Hua Wei (Pennsylvania State University); Chacha Chen (Shanghai Jiao Tong Univerisity), Guanjie Zheng (Pennsylvania State University); Kan Wu (Pennsylvania State University); Vikash Gayah (Pennsylvania State University); Kai Xu (Shanghai Tianrang Intelligent Technology Co., Ltd), Zhenhui Li (Pennsylvania State University)

- Unifying Inter-region Autocorrelation and Intra-region Structures for Spatial Embedding via Collective Adversarial Learning

2:15 PM - 2:20PM

Yunchao Zhang (Missouri University of Science and Technology);Pengyang Wang (Missouri University of Science and Technology);Xiaolin Li (Nanjing University);Yu Zheng (JD);Yanjie Fu (Missouri University of Science and Technology);

- Urban Traffic Prediction from Spatio-Temporal Data using Deep Meta Learning

2:20 PM - 2:25PM

Zheyi Pan (Shanghai Jiao Tong University);Yuxuan Liang (National University of Singapore);Weifeng Wang (Shanghai Jiao Tong University);Yong Yu (Shanghai Jiao Tong University);Yu Zheng (JD);Junbo Zhang (JD);

- UrbanFM: Inferring Fine-Grained Urban Flows

2:25 PM - 2:30PM

Yuxuan Liang (National University of Singapore); Kun Ouyang (National University of Singapore); Lin Jing (Xidian University); Sijie Ruan (Xidian University); Ye Liu (National University of Singapore); Junbo Zhang (JD); David Rosenblum (National University of Singapore); Yu Zheng (JD)

- Travel Time Estimation using Sparse Data

2:30 PM - 2:35PM

Nikolas Zygouras, Nikos Panagiotou, Yang Li, Leonidas Guibas, Dimitrios Gunopulos (University of Athens).