Vehicle Trajectories Classification using Kernel Evidential C-means algorithm for Failure Trajectory Prediction
EL BENDADI ; LAKHDAR ; SBAI ; KOITA
Type de document
COMMUNICATION AVEC ACTES INTERNATIONAL (ACTI)
Langue
anglais
Auteur
EL BENDADI ; LAKHDAR ; SBAI ; KOITA
Résumé / Abstract
In this article, we introduce a kernel version of the evidential C-means algorithm (KECM). The latter is obtained from a formulation of evidential C-means algorithm (ECM). The approach is applied to the classification of experimental data collected from a system called Vehicle-Infrastructure Driver (VID), based on several representative trajectories observations made in a bend. The test on real experimental data shows the value of the exploratory analysis method of data. In this work, the focus will be on the calculation of the intra-class variance to obtain adequate results with data experimentally realized based on the instructions given to drivers.
