基于睡眠传感器的体动异常标记与得分数据

基于睡眠传感器采集的用户睡眠数据，主要考察用户每5分钟的体动情况，把出现体动过多的时间点标记出来，并最终得到当晚的体动得分。用户可根据此数据，知晓当晚睡眠情况的好坏，给予用户可度量的正向反馈。此数据不仅能够为睡眠健康行业提供重要的参考信息，而且有助于建立行业标准，从而推动整个领域的发展和创新。1、收集智能床传感器的夜间体动次数的数据。这些数据经过异常值处理后，设置一个时间窗口5分钟，统计该5分钟的体动次数；
2、对体动次数进行异常标记的判断,成年人在夜间睡眠时每小时的体动次数正常值都在20次以下，因此设置每5分钟体动次数大于3次为体动异常，标记为1。
3、统计体动异常标记次数占总的夜间体动次数的占比，小于等于10%为5分，10%-15%为4分,15%-20%为3分，20%-25%为2分，25%-30%为1分， 大于30%为0分，作为该用户的当晚体动得分。

This dataset comprises user sleep data collected via sleep sensors, focusing on the user's body movement status every 5 minutes. Time points with excessive body movement are flagged, and a final nightly body movement score is generated. Users can use this data to evaluate their sleep quality for the night and obtain measurable positive feedback. This data not only provides critical reference information for the sleep health industry but also helps establish industry standards, thereby promoting the development and innovation of the entire field.
1. Collect data on nighttime body movement counts from smart bed sensors. After outlier handling, a 5-minute time window is set to count the number of body movements within this period.
2. Perform anomaly marking judgment on the body movement count. For adults, the normal number of nighttime body movements per hour is below 20. Therefore, body movement counts exceeding 3 times per 5 minutes are defined as abnormal and marked as 1.
3. Calculate the proportion of abnormally marked body movements relative to the total nighttime body movement count. The scoring criteria are as follows: 5 points if the proportion is ≤10%, 4 points for 10%–15%, 3 points for 15%–20%, 2 points for 20%–25%, 1 point for 25%–30%, and 0 points if the proportion exceeds 30%. This serves as the user's nightly body movement score.

搜集汇总

数据集介绍

特点

该数据集包含基于睡眠传感器采集的用户夜间体动数据，通过算法标记异常体动并计算得分，用于评估睡眠质量，适用于睡眠健康行业。