Index: Riebler 2016 - BYM2 Intuitive Spatial Model

Riebler 2016 - BYM2 Intuitive Spatial Model

Routing Summary

Full capture of Riebler, Sorbye, Simpson and Rue (2016), An intuitive Bayesian spatial model for disease mapping that accounts for scaling (SMMR; arXiv:1601.01180): the BYM2 parameterisation, graph-dependent scaling, PC priors for $({\tau }_b,\varphi )$, simulation and Sardinia IDDM application.

• Need the paper on one page, PyMC example E18, section map? → Riebler 2016 - BYM2 Intuitive Spatial Model - Overview
• Need Besag, BYM, Leroux, Dean definitions? → Riebler 2016 - Classical Disease Mapping Models Besag BYM Leroux Dean
• Need scaling factor, BYM2 definition, sparse form, PyMC sketch? → Riebler 2016 - Scaling and the BYM2 Construction
• Need PC priors and KLD derivation? → Riebler 2016 - Penalised Complexity Priors for BYM2
• Need Tables 1-3, simulation design, Sardinia results? → Riebler 2016 - Simulation Study and Sardinia IDDM Application

Concept Map

Concept	Note	Type	Depends On	Key Result
Besag/BYM/Leroux/Dean	Riebler 2016 - Classical Disease Mapping Models Besag BYM Leroux Dean	definition	none	Leroux cannot be scaled
Scaling, BYM2	Riebler 2016 - Scaling and the BYM2 Construction	theorem	classical models	${\sigma }_{GV}^2$ varies 0.29 to 4.78 across maps
PC priors	Riebler 2016 - Penalised Complexity Priors for BYM2	theorem	BYM2	$\theta =-\ln \alpha /U$; type-2 Gumbel
Simulation, Sardinia	Riebler 2016 - Simulation Study and Sardinia IDDM Application	example	PC priors	shrinks to both base models

Notes

• Riebler 2016 - BYM2 Intuitive Spatial Model - Overview — CONTAINS: question, contributions, E18 link, Source Map.
• Riebler 2016 - Classical Disease Mapping Models Besag BYM Leroux Dean — CONTAINS: Poisson setup, four latent models.
• Riebler 2016 - Scaling and the BYM2 Construction — CONTAINS: generalised variance, BYM2, sparse form, scaling code, PyMC sketch.
• Riebler 2016 - Penalised Complexity Priors for BYM2 — CONTAINS: PC principles, ${\tau }_b$ and $\varphi$ priors, KLD derivation, code.
• Riebler 2016 - Simulation Study and Sardinia IDDM Application — CONTAINS: Tables 1-3, discussion, INLA call.

Sources

Riebler 2016 - BYM2 An Intuitive Bayesian Spatial Model for Disease Mapping

See Also

Spatial Models - BYM | Simpson 2016 - Going Off Grid - Overview