Review & Perspective for Distance Based Trajectory Clustering

Philippe Besse (1, 2) , Brendan Guillouet (2) , Jean-Michel Loubes (2) , Royer François (3)

1 INSA Toulouse - Institut National des Sciences Appliquées - Toulouse
2 IMT - Institut de Mathématiques de Toulouse UMR5219
3 Datasio

In this paper we tackle the issue of clustering trajectories of geolocalized observations. Using clustering technics based on the choice of a distance between the observations, we first provide a comprehensive review of the different distances used in the literature to compare trajectories. Then based on the limitations of these methods, we introduce a new distance : Symmetrized Segment-Path Distance (SSPD). We finally compare this new distance to the others according to their corresponding clustering results obtained using both hierarchical clustering and affinity propagation methods.

Mots clés: Trajectory Clustering

Domaines: Machine Learning [stat.ML]; Applications [stat.AP]

Philippe Besse, Brendan Guillouet, Jean-Michel Loubes, Royer François. Review & Perspective for Distance Based Trajectory Clustering. 2015. ⟨hal-01184637⟩