Using machine learning techniques to characterize sleep-deprived driving behavior

Objective Sleep deprivation is known to affect driving behavior and may lead serious car accidents similar the effects from e.g., alcohol. In a previous study, we have demonstrated that use of machine learning techniques allows adequate characterization abnormal after alprazolam and/or alcohol intake. present extend this approach sleep test model for new interventions. We aimed classify deprivation, and, by using model, tested if could also pick up resulting other Methods Data were collected during in which 24 subjects being sleep-deprived well-rested night. Features calculated several parameters, such as lateral position, speed car, steering speed. used gradient boosting deprivation. The was validated 5-fold cross validation technique. Next, probability scores identify overlap affected current study alprazolam, alcohol, placebo are test/validate approach. Results detected simulator with an accuracy 77 ± 9%. Abnormal lesser extent intake, showed remarkably characteristics average score measurements 0.79, 0.63, only 0.27 0.30, matching expected relative drowsiness. Conclusion developed detecting induced shows similarities between interventions, i.e., alprazolam. Consequently, our serve next reference point battery newly drugs.