A generalized measure of uncertainty in geostatistical model selection

Chun-Shu Chen, Jun Zhu, Tingjin Chu

Research output: Contribution to journalReview article

1 Citation (Scopus)

Abstract

Model selection and model averaging are essential to regression analysis in environmental studies, but determining which of the two approaches is the more appropriate and under what circumstances remains an active research topic. In this paper, we focus on geostatistical regression models for spatially referenced environmental data. For a general information criterion, we develop a new perturbation-based criterion that measures the uncertainty (or, instability) of spatial model selection, as well as an empirical rule for choosing between model selection and model averaging. Statistical inference based on the proposed model selection instability measure is justified both in theory and via a simulation study. The predictive performance of model selection and model averaging can be quite different when the uncertainty in model selection is relatively large, but the performance becomes more comparable as this uncertainty decreases. For illustration, a precipitation data set in the state of Colorado is analyzed.

Original languageEnglish
Pages (from-to)203-228
Number of pages26
JournalStatistica Sinica
Volume28
Issue number1
DOIs
Publication statusPublished - 2018 Jan 1

Fingerprint

Model Selection
Model Averaging
Uncertainty
Information Criterion
Spatial Model
Statistical Inference
Regression Analysis
Model selection
Regression Model
Simulation Study
Perturbation
Decrease
Model averaging

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Chen, Chun-Shu ; Zhu, Jun ; Chu, Tingjin. / A generalized measure of uncertainty in geostatistical model selection. In: Statistica Sinica. 2018 ; Vol. 28, No. 1. pp. 203-228.
@article{f996cb6148a449da8eb2316ca96c21bf,
title = "A generalized measure of uncertainty in geostatistical model selection",
abstract = "Model selection and model averaging are essential to regression analysis in environmental studies, but determining which of the two approaches is the more appropriate and under what circumstances remains an active research topic. In this paper, we focus on geostatistical regression models for spatially referenced environmental data. For a general information criterion, we develop a new perturbation-based criterion that measures the uncertainty (or, instability) of spatial model selection, as well as an empirical rule for choosing between model selection and model averaging. Statistical inference based on the proposed model selection instability measure is justified both in theory and via a simulation study. The predictive performance of model selection and model averaging can be quite different when the uncertainty in model selection is relatively large, but the performance becomes more comparable as this uncertainty decreases. For illustration, a precipitation data set in the state of Colorado is analyzed.",
author = "Chun-Shu Chen and Jun Zhu and Tingjin Chu",
year = "2018",
month = "1",
day = "1",
doi = "10.5705/ss.202016.0368",
language = "English",
volume = "28",
pages = "203--228",
journal = "Statistica Sinica",
issn = "1017-0405",
publisher = "Institute of Statistical Science",
number = "1",

}

A generalized measure of uncertainty in geostatistical model selection. / Chen, Chun-Shu; Zhu, Jun; Chu, Tingjin.

In: Statistica Sinica, Vol. 28, No. 1, 01.01.2018, p. 203-228.

Research output: Contribution to journalReview article

TY - JOUR

T1 - A generalized measure of uncertainty in geostatistical model selection

AU - Chen, Chun-Shu

AU - Zhu, Jun

AU - Chu, Tingjin

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Model selection and model averaging are essential to regression analysis in environmental studies, but determining which of the two approaches is the more appropriate and under what circumstances remains an active research topic. In this paper, we focus on geostatistical regression models for spatially referenced environmental data. For a general information criterion, we develop a new perturbation-based criterion that measures the uncertainty (or, instability) of spatial model selection, as well as an empirical rule for choosing between model selection and model averaging. Statistical inference based on the proposed model selection instability measure is justified both in theory and via a simulation study. The predictive performance of model selection and model averaging can be quite different when the uncertainty in model selection is relatively large, but the performance becomes more comparable as this uncertainty decreases. For illustration, a precipitation data set in the state of Colorado is analyzed.

AB - Model selection and model averaging are essential to regression analysis in environmental studies, but determining which of the two approaches is the more appropriate and under what circumstances remains an active research topic. In this paper, we focus on geostatistical regression models for spatially referenced environmental data. For a general information criterion, we develop a new perturbation-based criterion that measures the uncertainty (or, instability) of spatial model selection, as well as an empirical rule for choosing between model selection and model averaging. Statistical inference based on the proposed model selection instability measure is justified both in theory and via a simulation study. The predictive performance of model selection and model averaging can be quite different when the uncertainty in model selection is relatively large, but the performance becomes more comparable as this uncertainty decreases. For illustration, a precipitation data set in the state of Colorado is analyzed.

UR - http://www.scopus.com/inward/record.url?scp=85040098506&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85040098506&partnerID=8YFLogxK

U2 - 10.5705/ss.202016.0368

DO - 10.5705/ss.202016.0368

M3 - Review article

AN - SCOPUS:85040098506

VL - 28

SP - 203

EP - 228

JO - Statistica Sinica

JF - Statistica Sinica

SN - 1017-0405

IS - 1

ER -