Abstract:We propose nonparametric methods for functional linear regression which are designed for sparse longitudinal data, where both the predictor and response are functions of a covariate such as time. Predictor and response processes have smooth random trajectories, and the data consist of a small number of noisy repeated measurements made at irregular times for a sample of subjects. In longitudinal studies, the number of repeated measurements per subject is often small and may be modeled as a discrete random number and, accordingly, only a finite and asymptotically nonincreasing number of measurements are available for each subject or experimental unit. We propose a functional regression approach for this situation, using functional principal component analysis, where we estimate the functional principal component scores through conditional expectations. This allows the prediction of an unobserved response trajectory from sparse measurements of a predictor trajectory. The resulting technique is flexible and allows for different patterns regarding the timing of the measurements obtained for predictor and response trajectories. Asymptotic properties for a sample of n subjects are investigated under mild conditions, as n →∞, and we obtain consistent estimation for the regression function. Besides convergence results for the components of functional linear regression, such as the regression parameter function, we construct asymptotic pointwise confidence bands for the predicted trajectories. A functional coefficient of determination as a measure of the variance explained by the functional regression model is introduced, extending the standard R to the functional case. The proposed methods are illustrated with a simulation study, longitudinal primary biliary liver cirrhosis data and an analysis of

On the functional regression model and its finite-dimensional approximations

A Bayesian approach to functional regression: theory and computation

Functional Linear Regression with Mixed Predictors

On functional logistic regression: some conceptual issues

An RKHS model for variable selection in functional regression

Partially Functional Linear Regression in High Dimensions

Varying-coefficient functional linear regression

Prediction in functional regression with discretely observed and noisy covariates

On estimation of the effect lag of predictors and prediction in functional linear model

Functional linear regression with points of impact

Robust Functional Regression with Discretely Sampled Predictors

Estimating the conditional distribution in functional regression problems

Variable Selection and Minimax Prediction in High-dimensional Functional Linear Model

Hybrid Regularisation of Functional Linear Models

Regression with Functional Errors-in-Predictors: A Generalized Method-of-Moments Approach

Estimation in Functional Single-Index Varying Coefficient Model

Scalar-on-function local linear regression and beyond

Partial functional linear regression with autoregressive errors

Partially functional linear varying coefficient model

Functional Linear Regression Analysis for Longitudinal Data 1 This Reprint Differs from the Original in Pagination and Typographic Detail. 1 2