Detection of non-recurrent road traffic events based on clustering indicators

Pierre-Antoine Laharotte (1) , Romain Billot (2, 3) , Nour Eddin El Faouzi (4)

1 DMOB - Département Mobilité (CEREMA)
2 Lab-STICC_IMTA_CID_DECIDE
3 IMT Atlantique - LUSSI - Département Logique des Usages, Sciences sociales et Sciences de l'Information
4 LICIT UMR TE - Laboratoire d'Ingénierie Circulation Transport

Based on a clustering indicator, an alteration of the classical road traffic indicators is proposed for incident detection. The resulting filter method reduces the inaccuracies of comparable detection method and enables to better separate usual traffic patterns from non-recurrent situations. Three alternative detection approaches are considered as baseline comparison for performance estimation.

Mots clés: TRAFFIC ENGINEERING; CLUSTERING; STATISTICAL ANALYSIS

Mots clés: CIRCULATION ROUTIERE; ANALYSE STATISTIQUE

Domaines: Ordinateur et société [cs.CY]; Applications [stat.AP]

Pierre-Antoine Laharotte, Romain Billot, Nour Eddin El Faouzi. Detection of non-recurrent road traffic events based on clustering indicators. ESANN'17 : European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, Apr 2017, Bruges, Belgium. ⟨hal-01802649⟩