Sleep stage estimation by learning classifier system towards care support for aged persons

This paper proposes the sleep stage estimation method that can provide more accurate estimation than the conventional method and is robust to bad condition of humans without connecting any devices to human's body. Our proposed method can extract the specific wave pattern required to estimate the sleep stage from heart beat data. Through the intensive simulations by using the actual data of the human subjects, the following implications have been revealed: (1) the proposed method can provide more accurate sleep stage estimation than the conventional methods, and (2) he sleep stage estimation is robust to the bad physical conditions.