论文标题
部分可观测时空混沌系统的无模型预测
Temporal dynamics of subjective sleepiness: A convergence analysis of two scales
论文作者
论文摘要
储层计算是预测湍流的有力工具，其简单的架构具有处理大型系统的计算效率。然而，其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间，然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用，该系统模拟了湍流的若干特征。我们表明，使用径向基函数作为非线性投影器，即使只有部分观测并且不知道控制方程，也能稳健地捕捉复杂的系统非线性。最后，我们表明，当测量稀疏、不完整且带有噪声，甚至控制方程变得不准确时，我们的网络仍然可以产生相当准确的预测，从而为实际湍流系统的无模型预测铺平了道路。
While sleepiness assessment metrics were initially developed in medical research to study the effects of drugs on sleep, subjective sleepiness assessment is now widely used in both fundamental and applied studies. The Stanford Sleepiness Scale (SSS) and the Karolinska Sleepiness Scale (KSS) are often considered the gold standard in sleepiness research. However, only a few studies have applied both scales, and their convergence and specific features have not been sufficiently investigated. The present study aims to analyse the dynamics and convergence of subjective sleepiness as measured by the KSS and SSS in a population of adults. To achieve this, we present the Subjective Sleepiness Dynamics Dataset (SSDD), which collects evening and morning data on situational subjective sleepiness. A total of 208 adults participated in the experiment. Our findings suggest that sleepiness generally increased from the evening till night and was highest early in the morning. The SSS score appeared to be more sensitive to certain factors, such as the presence of a sleep disorder. The SSS and KSS scores strongly correlated with each other and converged on sleepiness assessment. However, the KSS showed a more even distribution of scores than the SSS. Currently, we are continuously expanding the SSDD.
