A comparison of statistical methods for estimating individual location densities from smartphone data

In this work, the focus is on location data collected by smartphone applications. Specically, we propose and compare a set of models of increasing complexity to estimate individual location at any time, uncertainty included. Unlike classic tracking for high spatio-temporal resolution data, the approaches are suitable when location data are sparse in time and are affected by non negligible errors. The approaches build upon mixtures of densities that describe past and future locations; the model parameters are estimated by maximum likelihood. The approaches are applied to smartphone location data collected by the Earthquake Network citizen science project.