Predicting long-term sleep deprivation using wearable sensors and health surveys

Rafael Trujillo, Enshi Zhang, John Michael Templeton, Christian Poellabauer

Abstract: Highlights•Fitbit data was used with health surveys to predict long-term insufficient sleep.•Recursive Feature Elimination was used to select features.•Models were trained that included and excluded previous months’ sleep measurements.•The most important features were physical activity, past sleep, and depression.•Sex differences in feature selection and performance were analyzed.