Urban Mobility Safety in Barcelona: An Evaluation of Risk Factors and Techniques to Achieve Vision Zero Seminar

In order to converge towards the Vision Zero concept in road safety, traffic crashes require extensive study and analysis. This research is aimed at this goal, applied to the urban and metropolitan environment of Barcelona. We use different techniques according to the objectives and data: traditional statistical models, time series analysis, and machine learning techniques. Random Utility Models such as Binary Probit Model, Chi-square Automatic Interaction Detector, Bayesian networks such as Tree Augmented Naïve tree, and Kohonen clustering are all utilized. The methodology is generic and can be extrapolated to other cities, regions/countries, and datasets.