Trajectory similarity metric based on vector fields on manifold and multi-scale space partition
Heng Ma, Meng Wang · 2016
For trajectory model to study mining, using Vector Fields on Manifold instead of the Euclidean distance to metric similarity between trajectories, multi scale transform method is used to optimize the mapping in the Vector Fields on Manifold trajectory distance calculation and use Som algorithm for training a classification model. This method will be the trajectory shape features to measure the similarity between trajectories, multiscale partition reduces the Vector Fields on Manifold similarity measurement noise, Som, and solve the artificially set the number of clusters by clustering is not accurate. By calculating the Euclidean distance and trajectory point minimum distance respectively using Kmeans and Som as a comparative experiment. The results show that the method in the measurement of trajectory similarity accuracy is higher than that of the contrast experiment of the method using 1000 hurricane trajectory data.