Comparing Bayesian spatial models: Goodness-of-smoothing criteria for assessing under- and over-smoothing.
Journal
PloS one
ISSN: 1932-6203
Titre abrégé: PLoS One
Pays: United States
ID NLM: 101285081
Informations de publication
Date de publication:
2020
2020
Historique:
received:
26
02
2020
accepted:
27
04
2020
entrez:
21
5
2020
pubmed:
21
5
2020
medline:
1
8
2020
Statut:
epublish
Résumé
Many methods of spatial smoothing have been developed, for both point data as well as areal data. In Bayesian spatial models, this is achieved by purposefully designed prior(s) or smoothing functions which smooth estimates towards a local or global mean. Smoothing is important for several reasons, not least of all because it increases predictive robustness and reduces uncertainty of the estimates. Despite the benefits of smoothing, this attribute is all but ignored when it comes to model selection. Traditional goodness-of-fit measures focus on model fit and model parsimony, but neglect "goodness-of-smoothing", and are therefore not necessarily good indicators of model performance. Comparing spatial models while taking into account the degree of spatial smoothing is not straightforward because smoothing and model fit can be viewed as opposing goals. Over- and under-smoothing of spatial data are genuine concerns, but have received very little attention in the literature. This paper demonstrates the problem with spatial model selection based solely on goodness-of-fit by proposing several methods for quantifying the degree of smoothing. Several commonly used spatial models are fit to real data, and subsequently compared using the goodness-of-fit and goodness-of-smoothing statistics. The proposed goodness-of-smoothing statistics show substantial agreement in the task of model selection, and tend to avoid models that over- or under-smooth. Conversely, the traditional goodness-of-fit criteria often don't agree, and can lead to poor model choice. In particular, the well-known deviance information criterion tended to select under-smoothed models. Some of the goodness-of-smoothing methods may be improved with modifications and better guidelines for their interpretation. However, these proposed goodness-of-smoothing methods offer researchers a solution to spatial model selection which is easy to implement. Moreover, they highlight the danger in relying on goodness-of-fit measures when comparing spatial models.
Sections du résumé
BACKGROUND
Many methods of spatial smoothing have been developed, for both point data as well as areal data. In Bayesian spatial models, this is achieved by purposefully designed prior(s) or smoothing functions which smooth estimates towards a local or global mean. Smoothing is important for several reasons, not least of all because it increases predictive robustness and reduces uncertainty of the estimates. Despite the benefits of smoothing, this attribute is all but ignored when it comes to model selection. Traditional goodness-of-fit measures focus on model fit and model parsimony, but neglect "goodness-of-smoothing", and are therefore not necessarily good indicators of model performance. Comparing spatial models while taking into account the degree of spatial smoothing is not straightforward because smoothing and model fit can be viewed as opposing goals. Over- and under-smoothing of spatial data are genuine concerns, but have received very little attention in the literature.
METHODS
This paper demonstrates the problem with spatial model selection based solely on goodness-of-fit by proposing several methods for quantifying the degree of smoothing. Several commonly used spatial models are fit to real data, and subsequently compared using the goodness-of-fit and goodness-of-smoothing statistics.
RESULTS
The proposed goodness-of-smoothing statistics show substantial agreement in the task of model selection, and tend to avoid models that over- or under-smooth. Conversely, the traditional goodness-of-fit criteria often don't agree, and can lead to poor model choice. In particular, the well-known deviance information criterion tended to select under-smoothed models.
CONCLUSIONS
Some of the goodness-of-smoothing methods may be improved with modifications and better guidelines for their interpretation. However, these proposed goodness-of-smoothing methods offer researchers a solution to spatial model selection which is easy to implement. Moreover, they highlight the danger in relying on goodness-of-fit measures when comparing spatial models.
Identifiants
pubmed: 32433653
doi: 10.1371/journal.pone.0233019
pii: PONE-D-20-05534
pmc: PMC7239453
doi:
Types de publication
Comparative Study
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
e0233019Déclaration de conflit d'intérêts
The authors have declared that no competing interests exist.
Références
Accid Anal Prev. 2020 Feb;135:105323
pubmed: 31648775
Environmetrics. 2015 Sep;26(6):431-441
pubmed: 26900339
Biometrika. 1950 Jun;37(1-2):17-23
pubmed: 15420245
Spat Spatiotemporal Epidemiol. 2017 Feb;20:27-34
pubmed: 28137675
Phys Ther. 2005 Mar;85(3):257-68
pubmed: 15733050
Bayesian Anal. 2015 Dec 1;10(4):965-990
pubmed: 26753014
Biometrics. 1977 Mar;33(1):159-74
pubmed: 843571
Biometrics. 1987 Sep;43(3):671-81
pubmed: 3663823
AIMS Public Health. 2016 Mar 4;3(1):65-82
pubmed: 29546147
Stat Methods Med Res. 2005 Feb;14(1):35-59
pubmed: 15690999
Stat Med. 2002 Feb 15;21(3):359-70
pubmed: 11813223
Biometrics. 2000 Mar;56(1):13-21
pubmed: 10783772
Sci Total Environ. 2019 Feb 20;652:1339-1346
pubmed: 30586819
Accid Anal Prev. 2017 Mar;100:37-43
pubmed: 28088033
Int J Health Geogr. 2017 Dec 16;16(1):47
pubmed: 29246157
Int J Health Geogr. 2007 Nov 29;6:54
pubmed: 18045503