Séminaire R-D : High-Resolution Road Vehicle Collision Prediction for the City of Montréal

Séminaire R-D : High-Resolution Road Vehicle Collision Prediction for the City of Montréal
7/11/19 11h00
CRIM (405, avenue Ogilvy, bureau 101, Montréal)

Présentation en anglais

Prévision à haute résolution des collisions de véhicules routiers pour la Ville de Montréal


Tristan Glatardprofesseur agrégé au département d'informatique et de génie logiciel de l'Université Concordia, titulaire de la Chaire de recherche du Canada (Niveau II) sur les infrastructures de données massives pour la neuroinformatique et professeur adjoint au département d'informatique de l'Université McGill.

High-Resolution Road Vehicle Collision Prediction for the City of Montréal


Tristan Glatard, Associate Professor at Concordia University’s Department of Computer-Science and Software Engineering, Canada Research Chair (Tier II) on Big Data Infrastructures for Neuroinformatics, and Adjunct Professor at the School of Computer-Science at McGill.


Road accidents are an important issue in our modern societies, responsible for millions of deaths and injuries around the world. In Québec only, road accidents cause hundreds of deaths and tens of thousands of injuries every year. In this presentation, we show how one can leverage open datasets of a city like Montréal to create high-resolution accident prediction models, using state-of-the-art big data analytics.

Compared to other studies in road accident prediction, we have a much higher prediction resolution: our models predict the occurrence of an accident within an hour, on road segments defined by intersections. Such models could be used in the context of road accident prevention, but also to identify key factors that can lead to an accident and consequently help elaborate new policies.

Experimental results show that 85% of road vehicle collisions are detected by our model with a false positive rate of 13%. The examples identified as positive are likely to correspond to high risk situations.

In addition, we were able to identify the most important factors for predicting vehicle collisions for the area of Montréal: the number of accidents on the same road segment during previous years, the temperature, the day of the year, the hour and the level of visibility.


Conférence gratuite. Inscription requise.


  • ????Vous connaissez le @cooperathon? En collaboration avec @mtlnewtech, 12 organisations vous lancent défis d’innovati… https://t.co/JebU85JvOk
  • @inocanada - Deuxième appel à candidatures QUANTINO Vous avez une entreprise innovante en démarrage oeuvrant en opt… https://t.co/78cOFuEZq4