Figure 8
From: Spatio-temporal techniques for user identification by means of GPS mobility data
Effect of training set size on the classification accuracy. We measure the average classification accuracy (y-axis) as we increase the number of sampled spatio-temporal points (x-axis) for the CenceMe dataset. We show how the results vary as we reduce the size of the observed traces in the CenceMe dataset to 20% (blue), 40% (red), 60% (green) and 80% (magenta) of the original size. The results show that our classification framework is robust with respect to the number of available observations.