Causal Modeling of Traffic Events
This research aims to use big data analytics to enable the prediction of future traffic states to provide Traffic Management Centers (TMCs) with enhanced awareness of network conditions and risks and an ability to proactively respond.
We develop probabilistic graphical models (e.g., Bayesian Networks) and other statistical learning models that learn causal, spatio-temporal dependence relations between network conditions and traffic states from large-scale, multi-source data to predict the future states given various road condition scenarios.
Related Papers:
- Kim, J., Wang, G., 2016. Diagnosis and Prediction of Traffic Congestion on Urban Road Networks Using Bayesian Networks. Transportation Research Record: Journal of the Transportation Research Board, 2595, 108–118.