Problem to be solved: to determine the REM sleep of the subject to be measured with no accuracy.When the rate of increase in the median value of the heartbeat for each first period of the subject is greater than a predetermined threshold, it is judged that the start of the first REM sleep period is started and when the rate of decrease in the median value of the heartbeat per the first period decreases, the end of the first REM sleep period is determined Set.And, the frequency of the body motion of the subject was analyzed by frequency analysis for the second period, and by the analysis of the frequency analyzed data, the decision tree to obtain the REM sleep was generated by machine learning, and the majority decision of the REM sleep which was obtained in the generated multiple decision trees It is determined that the second period is the second REM sleep period. Then, the REM sleep period is determined based on the result of the determination of the first and second REM sleep periods.Diagram【課題】被測定者に負担をかけずに、被測定者のレム睡眠を精度良く判定できるようにする。【解決手段】被測定者の第1の期間毎の心拍の中央値の増加率が、所定の閾値以上であるとき、第1のレム睡眠期間の開始と判定し、第1の期間毎の心拍の中央値の減少率が減少したとき、第1のレム睡眠期間の終了を判定する。また、被測定者の体動を第2の期間毎に周波数解析し、その周波数解析したデータの分析で、レム睡眠の判定を得る決定木を機械学習により複数生成し、生成した複数の決定木で得られたレム睡眠か否かの判定の多数決から、第2の期間が第2のレム睡眠期間であると判定する。そして、これらの第1及び第2のレム睡眠期間の判定を行った結果から、レム睡眠期間の確定処理を行う。【選択図】図9
