A hybrid self-attention deep learning framework for multivariate sleep stage classification

Sleep is a complicated biological process and plays an essential role in health. Sleep occurs in cycle and involves different sleep stages, helping restore functions of body and mind, such as immune, nervous, skeletal, and muscular systems [1]. Unhealthy lifestyles and work-related stress may lead to sleep disturbances, which has become one of the serious issues in modern societies. Sleep disorders not only cause a reduction in physical performance during the day, but have negative effects on cognitive functions [2]. Moreover, some psychological and neurological diseases can also deteriorate normal sleep patterns [3].