CMAT, Universidade do Minho
In this seminar, we will introduce a clustering approach, based on the theoretical framework of dynamic neural fields (DNF), to discover in an unsupervised manner natural grouping in a set of points taking into account their spatial and temporal characteristics. Results obtained using real-world GPS trajectories from vehicles will be used to show the feasibility of the proposed approach. It allows to identify simulataneously stop locations and the time spent in each one. The impact of the obtained results on systems that automatically learn drivers’ daily routines from GPS trajectories will be discussed. The research has been developed in the context of the Easy Ride project with the company BOSCH.
 Ferreira, F., Wojtak, W., Fernandes, C., Guimarães, P., Monteiro, S., Bicho, E., & Erlhagen, W. (2021). Dynamic Identification of Stop Locations from GPS Trajectories Based on Their Temporal and Spatial Characteristics. In International Conference on Artificial Neural Networks (pp. 347-359). Springer, Cham.
 Wojtak, W., Ferreira, F., Guimarães, P., Barbosa, P., Monteiro, S., Erlhagen, W., & Bicho, E. (2021). Towards Endowing Intelligent Cars with the Ability to Learn the Routines of Multiple Drivers: A Dynamic Neural Field Model. In International Conference on Computational Science and Its Applications (pp. 337-349). Springer, Cham.