Functional CAR Models for Large Spatially Correlated Functional Datasets

Lin Zhang, Veerabhadran Baladandayuthapani, Hongxiao Zhu, Keith A. Baggerly, Tadeusz Majewski, Bogdan A. Czerniak, Jeffrey S. Morris

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

We develop a functional conditional autoregressive (CAR) model for spatially correlated data for which functions are collected on areal units of a lattice. Our model performs functional response regression while accounting for spatial correlations with potentially nonseparable and nonstationary covariance structure, in both the space and functional domains. We show theoretically that our construction leads to a CAR model at each functional location, with spatial covariance parameters varying and borrowing strength across the functional domain. Using basis transformation strategies, the nonseparable spatial-functional model is computationally scalable to enormous functional datasets, generalizable to different basis functions, and can be used on functions defined on higher dimensional domains such as images. Through simulation studies, we demonstrate that accounting for the spatial correlation in our modeling leads to improved functional regression performance. Applied to a high-throughput spatially correlated copy number dataset, the model identifies genetic markers not identified by comparable methods that ignore spatial correlations. Supplementary materials for this article are available online.

Original languageEnglish (US)
Pages (from-to)772-786
Number of pages15
JournalJournal of the American Statistical Association
Volume111
Issue number514
DOIs
StatePublished - Apr 2 2016

Fingerprint

Conditional Model
Spatial Correlation
Autoregressive Model
Nonseparable
Regression
Correlated Data
Functional Model
Functional Response
Spatial Model
Covariance Structure
High Throughput
Basis Functions
High-dimensional
Simulation Study
Unit
Modeling
Model
Demonstrate
Spatial correlation
Autoregressive model

Keywords

  • Conditional autoregressive model
  • Functional data analysis
  • Functional regression
  • Spatial functional data
  • Whole-organ histology and genetic maps

Cite this

Zhang, L., Baladandayuthapani, V., Zhu, H., Baggerly, K. A., Majewski, T., Czerniak, B. A., & Morris, J. S. (2016). Functional CAR Models for Large Spatially Correlated Functional Datasets. Journal of the American Statistical Association, 111(514), 772-786. https://doi.org/10.1080/01621459.2015.1042581

Functional CAR Models for Large Spatially Correlated Functional Datasets. / Zhang, Lin; Baladandayuthapani, Veerabhadran; Zhu, Hongxiao; Baggerly, Keith A.; Majewski, Tadeusz; Czerniak, Bogdan A.; Morris, Jeffrey S.

In: Journal of the American Statistical Association, Vol. 111, No. 514, 02.04.2016, p. 772-786.

Research output: Contribution to journalArticle

Zhang, L, Baladandayuthapani, V, Zhu, H, Baggerly, KA, Majewski, T, Czerniak, BA & Morris, JS 2016, 'Functional CAR Models for Large Spatially Correlated Functional Datasets', Journal of the American Statistical Association, vol. 111, no. 514, pp. 772-786. https://doi.org/10.1080/01621459.2015.1042581
Zhang, Lin ; Baladandayuthapani, Veerabhadran ; Zhu, Hongxiao ; Baggerly, Keith A. ; Majewski, Tadeusz ; Czerniak, Bogdan A. ; Morris, Jeffrey S. / Functional CAR Models for Large Spatially Correlated Functional Datasets. In: Journal of the American Statistical Association. 2016 ; Vol. 111, No. 514. pp. 772-786.
@article{e42f6fdc23154bb9a9f9ff7a2154b09e,
title = "Functional CAR Models for Large Spatially Correlated Functional Datasets",
abstract = "We develop a functional conditional autoregressive (CAR) model for spatially correlated data for which functions are collected on areal units of a lattice. Our model performs functional response regression while accounting for spatial correlations with potentially nonseparable and nonstationary covariance structure, in both the space and functional domains. We show theoretically that our construction leads to a CAR model at each functional location, with spatial covariance parameters varying and borrowing strength across the functional domain. Using basis transformation strategies, the nonseparable spatial-functional model is computationally scalable to enormous functional datasets, generalizable to different basis functions, and can be used on functions defined on higher dimensional domains such as images. Through simulation studies, we demonstrate that accounting for the spatial correlation in our modeling leads to improved functional regression performance. Applied to a high-throughput spatially correlated copy number dataset, the model identifies genetic markers not identified by comparable methods that ignore spatial correlations. Supplementary materials for this article are available online.",
keywords = "Conditional autoregressive model, Functional data analysis, Functional regression, Spatial functional data, Whole-organ histology and genetic maps",
author = "Lin Zhang and Veerabhadran Baladandayuthapani and Hongxiao Zhu and Baggerly, {Keith A.} and Tadeusz Majewski and Czerniak, {Bogdan A.} and Morris, {Jeffrey S.}",
year = "2016",
month = "4",
day = "2",
doi = "10.1080/01621459.2015.1042581",
language = "English (US)",
volume = "111",
pages = "772--786",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "514",

}

TY - JOUR

T1 - Functional CAR Models for Large Spatially Correlated Functional Datasets

AU - Zhang, Lin

AU - Baladandayuthapani, Veerabhadran

AU - Zhu, Hongxiao

AU - Baggerly, Keith A.

AU - Majewski, Tadeusz

AU - Czerniak, Bogdan A.

AU - Morris, Jeffrey S.

PY - 2016/4/2

Y1 - 2016/4/2

N2 - We develop a functional conditional autoregressive (CAR) model for spatially correlated data for which functions are collected on areal units of a lattice. Our model performs functional response regression while accounting for spatial correlations with potentially nonseparable and nonstationary covariance structure, in both the space and functional domains. We show theoretically that our construction leads to a CAR model at each functional location, with spatial covariance parameters varying and borrowing strength across the functional domain. Using basis transformation strategies, the nonseparable spatial-functional model is computationally scalable to enormous functional datasets, generalizable to different basis functions, and can be used on functions defined on higher dimensional domains such as images. Through simulation studies, we demonstrate that accounting for the spatial correlation in our modeling leads to improved functional regression performance. Applied to a high-throughput spatially correlated copy number dataset, the model identifies genetic markers not identified by comparable methods that ignore spatial correlations. Supplementary materials for this article are available online.

AB - We develop a functional conditional autoregressive (CAR) model for spatially correlated data for which functions are collected on areal units of a lattice. Our model performs functional response regression while accounting for spatial correlations with potentially nonseparable and nonstationary covariance structure, in both the space and functional domains. We show theoretically that our construction leads to a CAR model at each functional location, with spatial covariance parameters varying and borrowing strength across the functional domain. Using basis transformation strategies, the nonseparable spatial-functional model is computationally scalable to enormous functional datasets, generalizable to different basis functions, and can be used on functions defined on higher dimensional domains such as images. Through simulation studies, we demonstrate that accounting for the spatial correlation in our modeling leads to improved functional regression performance. Applied to a high-throughput spatially correlated copy number dataset, the model identifies genetic markers not identified by comparable methods that ignore spatial correlations. Supplementary materials for this article are available online.

KW - Conditional autoregressive model

KW - Functional data analysis

KW - Functional regression

KW - Spatial functional data

KW - Whole-organ histology and genetic maps

UR - http://www.scopus.com/inward/record.url?scp=84983348732&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84983348732&partnerID=8YFLogxK

U2 - 10.1080/01621459.2015.1042581

DO - 10.1080/01621459.2015.1042581

M3 - Article

VL - 111

SP - 772

EP - 786

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 514

ER -