Title: Correcting selection bias via functional empirical Bayes
Authors: Yingying Fan - University of Southern California (United States) [presenting]
Abstract: Consider the problem of estimating mean curves in the setting of functional data, where each observed functional curve can be decomposed into a random mean curve and a error curve around the mean curve. Instead of using the naive method which simply estimates the mean curves using the observed curves, we propose a new method, Functional Empirical Bayes, to reduce the estimation bias. We theoretically study the proposed method and use real data sets to demonstrate the performance of it.