Individual sleep stage estimation and appropriate wakeup time judgment by learning classifier system for care support

为了减少养老院对老年人的夜间定期检查及其睡眠支持的严重问题，本研究提出了一种护理支持系统，该系统可以估计(1)针对每个老年人的总睡眠阶段估计；(2)对老年人状况较差的总睡眠阶段；(3)快速眼动睡眠阶段，用于判断老年人合适的起床时间。人体实验验证了所提出的护理支持系统的有效性。

To reduce a regular night round inspection of aged persons and their sleep support, which is the serious problem in care houses, this research proposes the care support system that can estimate (1) the total sleep stage estimation specialized to each aged person ; (2) the total sleep stage which is robust to the bad condition of aged persons ; and (3) the REM sleep stage for judging the appropriate wakeup time of aged persons. The human subject experiments validate the effectiveness of the proposed care support system.