MULTILEVEL FUNCTIONAL PRINCIPAL COMPONENT ANALYSIS

By Chong-Zhi Di, Ciprian M. Crainiceanu, Brian C. Caffo, Naresh M. Punjabi. Annals of Applied Statistics, 2009

388 citations recorded by the source corpus.

The Sleep Heart Health Study (SHHS) is a comprehensive landmark study of sleep and its impacts on health outcomes. A primary metric of the SHHS is the in-home polysomnogram, which includes two electroencephalographic (EEG) channels for each subject, at two visits. The volume and importance of this data presents enormous challenges for analysis. To address these challenges, we introduce multilevel functional principal component analysis (MFPCA), a novel statistical methodology designed to extract core intra- and inter-subject geometric components of multilevel functional data. Though motivated by the SHHS, the proposed methodology is generally applicable, with potential relevance to many modern scientific studies of hierarchical or longitudinal functional outcomes. Notably, using MFPCA, we identify and quantify associations between EEG activity during sleep and adverse cardiovascular outcomes.

Subfields: Cognitive Neuroscience, Food Science

View Paper: https://www.semanticscholar.org/paper/23aa60a2fe3e50347afccd462366532d386463dc

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Food Science

Citations in the corpus

Linguistic pitch analysis using functional principal component mixed effect models (2010, methodology): “Methods have recently been developed for hierarchical FPCA random-effects models (Di et al., 2009), but, because of the large number of covariates that are likely to affect the data, including emphasis on the modelling of random subject effects, neither the hierarchical nor the single-index…”
Multivariate Functional Principal Component Analysis for Data Observed on Different (Dimensional) Domains (2015, methodology): “CA u based on univariate FPCA for each element is computed using the PACE algorithm (Yao et al. 2005) with penalized splines to smooth the covariance function, as implemented in the R package refund (Di et al. 2009, Crainiceanu et al. 2014). The number of univariate principal components for each ele”
Penalized Functional Regression (2011, neutral): “Specifically, the method of Crainiceanu, Staicu, and Di (2009) uses MFPCA to construct a parsimonious basis for Xi(t) and uses a mixed model to estimate the PC loadings cik .”
Likelihood Ratio Tests for Dependent Data with Applications to Longitudinal and Functional Data Analysis (2014, methodology): “Detailed descriptions of this study can be found in Quan et al. (1997), Crainiceanu et al. (2009) and Di et al. (2009).”
A survey of functional principal component analysis (2014, background): “In such a case , Di et al .”
A general framework for functional regression modelling (2017, background): “Di et al. (2009), Greven et al. (2010), Shou et al. (2015) and Cederbaum et al. (2016) discuss the estimation of the eigenfunctions and eigenvalues for more complex functional random effects models.”
A Two Sample Distribution-Free Test for Functional Data with Application to a Diffusion Tensor Imaging Study of Multiple Sclerosis (2016, methodology): “Ghosh rodents of different sexes, Dorris et al. (2014) compare the distribution of each of several characteristics of interest across the two gender groups.”
Foundations of Linear and Generalized Linear Models (2015, methodology): “1 Longitudinal data: For overviews of GLMMs and marginal models, with emphasis on modeling longitudinal data, see Diggle et al. (2002), Fitzmaurice et al. (2011), and Hedeker and Gibbons (2006).”
Fast bivariate P‐splines: the sandwich smoother (2010, methodology): “P , ψ 4 } are the eigenfunctions from either of the following Case 1: The above two sets of eigenfunctions were used in Di et al. (2009), Greven et al. (2010), and Zipunnikov et al. (2011).”