Statistics > Applications
[Submitted on 11 Feb 2015 (v1), revised 11 Feb 2016 (this version, v2), latest version 17 Jan 2017 (v3)]
Title:Conditional Spectral Analysis of Replicated Multiple Time Series with Application to Nocturnal Physiology
View PDFAbstract:During sleep, the human body cycles through different states, and physiological activity during these states is essential to the rejuvenating properties of sleep. Researchers use polysomnography to record electrophysiological time series during sleep with the goal of characterizing sleep and elucidating the pathways through which sleep affects, and can be treated to improve, health and functioning. Important physiological information is contained in frequency patterns of many of these series, such as heart rate variability, so that the goal of an analysis of such data is to quantify associations between power spectra of time series from different sleep periods with cross-sectional clinical and behavioral variables. Such analyses are challenging as they must overcome the complex geometric structure and high dimensionality of spectral matrices. This article introduces a novel approach to analyzing such data by modeling Cholesky components of spectral matrices as bivariate functions of frequency and cross-sectional variables which, critically, preserves the poistive definite and Hermitian structure of spectral matrices while nonparametrically modeling associations. A fully automated Whittle likelihood MCMC algorithm enables inference on the association between the cross-sectional variable with the entire multivariate power spectrum and with any low-dimensional band-collapsed spectral measure. The approach not only provides inference on connections between clinical and behavioral variables and within-period spectra, but also presents the first methodology in the literature that can quantify connections with cross-period coherence, allowing for a deeper understanding of nocturnal physiology. The method is motivated by and used to analyze data from a study of poor sleep in older adults and finds a new link between autonomic nervous system activity and self-reported sleep.
Submission history
From: Robert Krafty [view email][v1] Wed, 11 Feb 2015 00:04:05 UTC (287 KB)
[v2] Thu, 11 Feb 2016 19:00:48 UTC (596 KB)
[v3] Tue, 17 Jan 2017 23:52:33 UTC (595 KB)
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.