Understanding human sleep behaviour by machine learning

Long term sleep quality assessment is essential to diagnose sleep disorders and to continuously monitor the health status. However, traditional polysomnography techniques are not suitable for long-term monitoring, whereas, methods able to continuously monitor the sleep pattern in an unobtrusive way are needed. In this paper, we present a general purpose sleep monitoring system that can be used for the pressure ulcer risk assessment, to monitor bed exits, and to observe the influence of medication on the sleep behaviour. Moreover, we compare several supervised learning algorithms in order to determine the most suitable in this context. Experimental results obtained by comparing the selected supervised algorithms show that we can accurately infer sleep duration, sleep positions, and routines with a completely unobtrusive approach.