Testing Separability of Functional Time Series

Panayiotis Constantinou, Piotr Kokoszka, Matthew Logan Reimherr

Research output: Contribution to journalArticle

Abstract

We derive and study a significance test for determining whether a panel of functional time series is separable. In the context of this paper, separability means that the covariance structure factors into the product of two functions, one depending only on time and the other depending only on the coordinates of the panel. Separability is a property that can dramatically improve computational efficiency by substantially reducing model complexity. It is especially useful for functional data, as it implies that the functional principal components are the same for each member of the panel. However, such an assumption must be verified before proceeding with further inference. Our approach is based on functional norm differences and provides a test with well-controlled size and high power. We establish our procedure quite generally, allowing one to test separability of autocovariances as well. In addition to an asymptotic justification, our methodology is validated by a simulation study. It is applied to functional panels of particulate pollution and stock market data.

Original languageEnglish (US)
Pages (from-to)731-747
Number of pages17
JournalJournal of Time Series Analysis
Volume39
Issue number5
DOIs
StatePublished - Sep 1 2018

Fingerprint

Separability
Computational efficiency
Time series
Pollution
Testing
Autocovariance
Significance Test
Functional Data
Model Complexity
Structure Factor
Covariance Structure
Principal Components
Stock Market
Justification
Computational Efficiency
High Power
Simulation Study
Norm
Imply
Methodology

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

Cite this

Constantinou, Panayiotis ; Kokoszka, Piotr ; Reimherr, Matthew Logan. / Testing Separability of Functional Time Series. In: Journal of Time Series Analysis. 2018 ; Vol. 39, No. 5. pp. 731-747.
@article{cf4a6954d1154d0aa27cb8e19672b478,
title = "Testing Separability of Functional Time Series",
abstract = "We derive and study a significance test for determining whether a panel of functional time series is separable. In the context of this paper, separability means that the covariance structure factors into the product of two functions, one depending only on time and the other depending only on the coordinates of the panel. Separability is a property that can dramatically improve computational efficiency by substantially reducing model complexity. It is especially useful for functional data, as it implies that the functional principal components are the same for each member of the panel. However, such an assumption must be verified before proceeding with further inference. Our approach is based on functional norm differences and provides a test with well-controlled size and high power. We establish our procedure quite generally, allowing one to test separability of autocovariances as well. In addition to an asymptotic justification, our methodology is validated by a simulation study. It is applied to functional panels of particulate pollution and stock market data.",
author = "Panayiotis Constantinou and Piotr Kokoszka and Reimherr, {Matthew Logan}",
year = "2018",
month = "9",
day = "1",
doi = "10.1111/jtsa.12302",
language = "English (US)",
volume = "39",
pages = "731--747",
journal = "Journal of Time Series Analysis",
issn = "0143-9782",
publisher = "Wiley-Blackwell",
number = "5",

}

Testing Separability of Functional Time Series. / Constantinou, Panayiotis; Kokoszka, Piotr; Reimherr, Matthew Logan.

In: Journal of Time Series Analysis, Vol. 39, No. 5, 01.09.2018, p. 731-747.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Testing Separability of Functional Time Series

AU - Constantinou, Panayiotis

AU - Kokoszka, Piotr

AU - Reimherr, Matthew Logan

PY - 2018/9/1

Y1 - 2018/9/1

N2 - We derive and study a significance test for determining whether a panel of functional time series is separable. In the context of this paper, separability means that the covariance structure factors into the product of two functions, one depending only on time and the other depending only on the coordinates of the panel. Separability is a property that can dramatically improve computational efficiency by substantially reducing model complexity. It is especially useful for functional data, as it implies that the functional principal components are the same for each member of the panel. However, such an assumption must be verified before proceeding with further inference. Our approach is based on functional norm differences and provides a test with well-controlled size and high power. We establish our procedure quite generally, allowing one to test separability of autocovariances as well. In addition to an asymptotic justification, our methodology is validated by a simulation study. It is applied to functional panels of particulate pollution and stock market data.

AB - We derive and study a significance test for determining whether a panel of functional time series is separable. In the context of this paper, separability means that the covariance structure factors into the product of two functions, one depending only on time and the other depending only on the coordinates of the panel. Separability is a property that can dramatically improve computational efficiency by substantially reducing model complexity. It is especially useful for functional data, as it implies that the functional principal components are the same for each member of the panel. However, such an assumption must be verified before proceeding with further inference. Our approach is based on functional norm differences and provides a test with well-controlled size and high power. We establish our procedure quite generally, allowing one to test separability of autocovariances as well. In addition to an asymptotic justification, our methodology is validated by a simulation study. It is applied to functional panels of particulate pollution and stock market data.

UR - http://www.scopus.com/inward/record.url?scp=85046339358&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046339358&partnerID=8YFLogxK

U2 - 10.1111/jtsa.12302

DO - 10.1111/jtsa.12302

M3 - Article

VL - 39

SP - 731

EP - 747

JO - Journal of Time Series Analysis

JF - Journal of Time Series Analysis

SN - 0143-9782

IS - 5

ER -