Abstract:We propose a nonparametric method to perform functional principal components analysis for the case of sparse longitudinal data. The method aims at irregularly spaced longitudinal data, where the number of repeated measurements available per subject is small. In contrast, classical functional data analysis requires a large number of regularly spaced measurements per subject. We assume that the repeated measurements are located randomly with a random number of repetitions for each subject and are determined by an underlying smooth random (subject-specific) trajectory plus measurement errors. Basic elements of our approach are the parsimonious estimation of the covariance structure and mean function of the trajectories, and the estimation of the variance of the measurement errors. The eigenfunction basis is estimated from the data, and functional principal components score estimates are obtained by a conditioning step. This conditional estimation method is conceptually simple and straightforward to implement. A key step is the derivation of asymptotic consistency and distribution results under mild conditions, using tools from functional analysis. Functional data analysis for sparse longitudinal data enables prediction of individual smooth trajectories even if only one or few measurements are available for a subject. Asymptotic pointwise and simultaneous confidence bands are obtained for predicted individual trajectories, based on asymptotic distributions, for simultaneous bands under the assumption of a finite number of components. Model selection techniques, such as the Akaike information criterion, are used to choose the model dimension corresponding to the number of eigenfunctions in the model. The methods are illustrated with a simulation study, longitudinal CD4 data for a sample of AIDS patients, and time-course gene expression data for the yeast cell cycle.

Effective dimension reduction for sparse functional data

Functional sufficient dimension reduction through distance covariance

On efficient dimension reduction with respect to a statistical functional of interest

Functional Sufficient Dimension Reduction for Functional Data Classification

Linear Dynamic Sparse Modelling for Functional MR Imaging.

Functional sufficient dimension reduction through information maximization with application to classification

Supervised dimension reduction for functional time series

A Functional approach for Two Way Dimension Reduction in Time Series

Dimension Reduction for Functional Regression with a Binary Response

Functional Data Analysis for Sparse Longitudinal Data

Partially Functional Linear Regression in High Dimensions

Locally Sparse Function on function Regression

Covariate Dependent Sparse Functional Data Analysis

FUNCTIONAL SUFFICIENT DIMENSION REDUCTION THROUGH AVERAGE FRÉCHET DERIVATIVES

Functional Linear Regression with Mixed Predictors

Inverse regression for spatially distributed functional data

Fast covariance estimation for multivariate sparse functional data

Dynamic Multivariate Functional Data Modeling Via Sparse Subspace Learning

Fast Covariance Estimation for Sparse Functional Data

Group sparse sufficient dimension reduction: a model-free group variable selection method

From sparse to dense functional data and beyond