An Iterative Strategy for Deep Learning Classification on Spatial Data Streams

Although the classification of a static set of spatial objects has been well studied, the classification of spatial objects generated by a spatial data stream has not. In this paper, we propose an iterative deep learning strategy for spatial data stream classification. Using a deep neural network, our strategy iteratively performs training and testing of the classifier, with the goal of reaching a desired accuracy, and the same accuracy that would be achieved as a classifier that is trained and tested with the entire object set. An experimental evaluation of our strategy versus a fully trained classifier with varying training and testing splits shows that the higher the percentage of training objects, the better the accuracy, although the other splits do converge closer to the accuracy of the fully trained classifier.