Table 1.

Summary of features extracted from preprocessed geolocation data.

Feature name	Description
Location variance	A measure of the variance in the location coordinates visited.
Number of clusters	The number of unique locations visited.
Entropy of locations	The information-theoretic entropy calculated on the proportion of time spent in each of the locations visited.
Normalized entropy	The entropy of locations feature normalized by dividing by the log of the number of location visited, resulting in a feature ranging between 0 and 1, which is less correlated with the number of clusters feature.
Home stay	The percentage of time that the individual is recorded at home.
Transition time	The percentage of time that the individual is recorded traveling between locations.
Total distance	The total distance traveled by the individual.
Diurnal movement	A measure of the diurnal regularity in the movements of the individual, calculated from the power in sinusoids fitted to the data with periods around 24 hours.
Diurnal movement on normalized coordinates	Similar to the diurnal movement feature, but calculated on normalized coordinates, making it less sensitive to the different distances that individuals may travel.
Diurnal movement on distance from home	Similar to the diurnal movement and diurnal movement on normalized coordinates features, but calculated on the single dimensional distance of the current location coordinates from the home location of the individual.