Paradiso S., Ziakopoulos A., Fortsakis P., Yannis G., “Probabilistic Modeling for Node-Based Partitioning of Telematics-Informed Road Networks” at the Proceedings of the 8th IRTAD International Conference, Athens, 15-17 April 2026.

The present study aims to explore driving behaviours mapped onto a road network, therefore studying a telematics-informed network, using a probabilistic machine learning model. A Gaussian Mixture Model (GMM) was employed in order to obtain a probabilistic node-based clustering to explore the structure of telematics-informed nodes within a study area. The analysis also involved the use of Transformer-based GNN to enhance the quality of node representations used as input for the GMM.