Predicting changes in wellbeing based on gait data
摘要
We report here the development of a technique for predicting a person’s wellbeing based on double-step accelerograms obtained using a smartphone while walking. Measurements of double steps with a smartphone were made with the device positioned without rigid fixation in the front pocket of lower body clothing. The value of the metric F1-score for the found structure of the vocal regressor was 0.86 for predicting lower back pain.