A0548
Title: Empirical likelihood inference for functional mean models with application to human cognitive impairment
Authors: Honglang Wang - Indiana University Indianapolis (United States) [presenting]
Xiang Wang - Indiana University Indianapolis (United States)
Abstract: The two-step-refining inference is considered for the mean function of sparse functional data to account for within-subject correlation. The refined estimator improves the efficiency of the local kernel smoothing estimator, which assumes a working independence correlation structure. The empirical likelihood (EL) based inference is proposed for the mean function of functional data with a bias-correlated estimating equation derived from the two-step-refining procedure. It not only establishes the asymptotic normality of the refined estimator but also derives Wilk's theorem for the empirical likelihood ratio test. The proposed methods perform favorably in finite sample applications from the simulation studies, as well as real data application to the Alzheimer's disease neuroimaging initiative (ADNI) study.
