Association Pattern Based Prediction for Moving Objects

This paper addresses the prediction for moving objects when the motions of the objects having some associations. In the real world, the moving objects often have some associations with other objects. Unfortunately, existing techniques for moving objects prediction are designed for single independent object, and ignore the associations between the moving objects. This paper firstly defines an association pattern based prediction model, which uses the association patterns of the moving objects to predict the trajectory of the moving objects, and then presents the Association Pattern based Prediction algorithm (APP algorithm) for moving objects. Theoretical analyses results show that, the APP algorithm has good efficiency for predicting moving objects.