Online Program Home
My Program

Abstract Details

Activity Number: 365
Type: Contributed
Date/Time: Tuesday, August 2, 2016 : 10:30 AM to 12:20 PM
Sponsor: Business and Economic Statistics Section
Abstract #318961
Title: Testing Separability of Functional Time Series
Author(s): Panayiotis Constantinou* and Piotr Kokoszka and Matthew Reimherr
Companies: Penn State University and Colorado State University and Penn State University
Keywords: Functional data analysis ; Separability ; Hypothesis Testing

In this work I will present a new statistical test for determining if a panel of functional time series is separable. Separability is property which can dramatically improve statistical efficiency while substantially reducing model complexity. In this context, separability means that the covariance structure factors into the produce of two functions, one depending only on time and the other depending only on the coordinates of the panel. Separability is especially useful for functional data as it means that functional principal components are the same for each member of the panel. However, in practice, such a strong assumption must be checked. Our test is based on functional norm differences and provides a very stable and powerful test. Our methodology will be illustrated via simulations and a financial application. Asymptotic theory will also be discussed.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association