I’d like to call your attention to a special session we’re organizing on "Transport Modeling with Machine Learning and Travel Behavior Methods”, in the IEEE Intelligent Transport Systems Conference 2018 (November 4-7, 2018, Maui, Hawaii, USA).
As proposed recently in the TRB ADB40 subcommittee (“Transportation Demand Forecast”), this is part of a strategy to bring together the two communities of Computer Science (particularly Machine Learning) and Travel Demand Modeling together. So we encourage everyone to join, if you come from any (or both) of these angles.
Here is the summary of the special session:
By its nature, Transportation is ultimately designed for people, be it for personal mobility or even for goods. Thus, understanding the demand side individually and collectively is the key motivation for decades of research in statistics, transport economics, behavior econometrics, transport psychology, among others. As a confluence of these, the area of travel behavior research summarizes much of the key findings and models in use around the world. More recently, due to increased data availability and computing power, machine learning techniques have started to become highly valuable tools to solve some of these “old” challenges together with new ones. On the other hand, many machine learning techniques are often accused of not being able to incorporate the dense domain knowledge accumulated in other areas, such as behavior econometrics.
This special session aims to bridge the gap between machine learning and econometrics tools, and will focus on topics such as:
- Deep Learning applied to behavior modeling
- Probabilistic graphical models applied to behavior modeling
- Data Sciences applied to behavior modeling
- Hybrid methods that combine ML and classical statistical methods
- Discrete choice modeling
- Structural Equations modeling
- Comparative studies (e.g. ML vs econometrics)
Also, please, pay attention to our website: https://www.ieee-itsc2018.org.
ML^2 - Machine Learning for Mobility Lab, Transport Modeling Division
Department of Management Engineering
Technical University of Denmark (DTU)