Premium
Vector Field k ‐Means: Clustering Trajectories by Fitting Multiple Vector Fields
Author(s) -
Ferreira Nivan,
Klosowski James T.,
Scheidegger Carlos E.,
Silva Cláudio T.
Publication year - 2013
Publication title -
computer graphics forum
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.578
H-Index - 120
eISSN - 1467-8659
pISSN - 0167-7055
DOI - 10.1111/cgf.12107
Subject(s) - cluster analysis , trajectory , vector field , computer science , field (mathematics) , similarity (geometry) , data mining , euclidean vector , cluster (spacecraft) , vector (molecular biology) , artificial intelligence , mathematics , image (mathematics) , physics , geometry , astronomy , pure mathematics , programming language , biochemistry , chemistry , gene , recombinant dna
Scientists study trajectory data to understand trends in movement patterns, such as human mobility for traffic analysis and urban planning. In this paper, we introduce a novel trajectory clustering technique whose central idea is to use vector fields to induce a notion of similarity between trajectories, letting the vector fields themselves define and represent each cluster. We present an efficient algorithm to find a locally optimal clustering of trajectories into vector fields, and demonstrate how vector‐field k‐means can find patterns missed by previous methods. We present experimental evidence of its effectiveness and efficiency using several datasets, including historical hurricane data, GPS tracks of people and vehicles, and anonymous cellular radio handoffs from a large service provider.