Department of Civil Engineering
Faculty of Engineering and Architecture Science
The use of learning approaches in modeling traffic
Date: Thursday August 16, 2018. 11:00AM—12:00PM
The presentation will tackle different examples of the use of learning approaches in the field of transportation, in particular in the case of transport modeling. First, the applications in the context of population synthesis and activity pattern generation are illustrated. Recent advances in agent-based micro-simulation modeling have further highlighted the importance of a thorough full synthetic population procedure for guaranteeing the correct characterization of real-world populations and underlying travel demands. In this regard, we propose an integrated approach including Markov Chain Monte Carlo (MCMC) simulation and profiling-based methods to capture the behavioral complexity and the great heterogeneity of agents of the true population through representative micro-samples. The population synthesis method is capable of building the joint distribution of a given population with its corresponding marginal distributions using either full or partial conditional probabilities or both of them simultaneously. In particular, the estimation of socio-demographic or transport-related variables and the characterization of daily activity-travel patterns are included within the framework. The fully probabilistic structure based on Markov Chains characterizing this framework makes it innovative compared to standard activity-based models. Moreover, data stemming from the 2010 Belgian Household Daily Travel Survey (BELDAM) are used to calibrate the modeling framework. We illustrate that this framework effectively captures the behavioral heterogeneity of travelers. Furthermore, we demonstrate that the proposed framework is adequately adapted to meeting the demand for large-scale micro-simulation scenarios of transportation and urban systems. Afterwards, a new Singular Value Decomposition-Latent Dirichlet Allocation (SVD-LDA) model will be presented to detect congestion patterns. The model is calibrated based on the taxi trip dataset, collected in the city of Chicago in 2016. The model specifies the trends in terms of traffic congestion patterns, also called topics in the literature. The main findings of this study reveal that, within day, SVD-LDA succeeds in distinguishing traffic patterns in terms of peak and off-peak hours with highly balanced topics. Furthermore, for day-to-day patterns, the model highlights stable traffic patterns across days. This lack of daily variability resulted in highly unbalanced share of topics.
Mario Cools (Lier, 1982) holds a master degree in applied economics, major quantitative business economics, minor operations research (University of Antwerp, 2004) and a master degree in applied statistics (Hasselt University, 2005). After obtaining his statistics degree, he started working as a PhD candidate at the Transportation Research Institute of Hasselt University, where he obtained his PhD in transportation sciences in 2009. After having continued his research as a post-doctoral researcher, Mario Cools was granted a post-doctoral fellowship from the Research Foundation Flanders (FWO) in 2010. Consequently, he worked at the KU Leuven Campus Brussels as an assistant professor in the Faculty of Economics and Management, and 10% as assistant professor the Faculty of Business Economics of Hasselt University. Currently, Mario Cools is appointed as full-time associate professor at the Faculty of Applied Sciences of the University of Liège. He is author and co-author of different scientific publications in research domains such as travel behaviour, transport policy and activity-based travel demand models.