A0296
Title: Spectral density estimation of function-valued spatial processes
Authors: Tailen Hsing - University of Michigan (United States) [presenting]
Stilian Stoev - University of Michigan (United States)
Rafail Kartsioukas - University of Michigan (United States)
Abstract: The spectral density of a stationary process fully characterizes the second- order properties of the process. We are interested in the estimation of the spectral density of a continuous-parameter stochastic process taking values in an infinite-dimensional Hilbert space. We assume that the process is observed at irregularly-spaced points, which is common in spatial statistics. We consider a lag-window estimator and explore the asymptotic properties of the estimator.
