Home
This Title All WIREs
WIREs RSS Feed
How to cite this WIREs title:
WIREs Comp Stat

Current Issue

Wiley Interdisciplinary Reviews:
Computational Statistics
Volume 10 Issue 6 (November/December 2018)
Page 0 - 0

Overview

Bayesian latent modeling of spatio‐temporal variation in small‐area health data
Published Online: Jun 29 2018
DOI: 10.1002/wics.1441
The analysis of small area health data is often a focus of epidemiological research. While it is possible to provide smoothed risk estimates for disease maps, it is often important to consider underlying structure in the risk outcome that is suggested by understanding of the etiology of disease processes. To address this more complex problem, it can be important to consider latent structure in the disease risk. Latent structure can take a variety of forms, from basic random effects to more complex latent variable models. In this paper I will review outline the basic approaches to the problem of latent structure for spatio‐temporal health outcome data and the solutions that Bayesian modeling can offer.
Abstract Full article on Wiley Online Library:   HTML | PDF

Advanced Reviews

Bayesian variable selection using the hyper‐ g prior in WinBUGS
Published Online: Jul 17 2018
DOI: 10.1002/wics.1442
Posterior inclusion probabilities using hyper‐g prior is a default choice for Bayesian variable selection (BVS). In this article we provide an overview of the BVS and explain in detail the specification for the hyper‐g prior setup using WinBUGS.
Abstract Full article on Wiley Online Library:   HTML | PDF
Spatial modeling with R‐INLA: A review
Published Online: Jul 05 2018
DOI: 10.1002/wics.1443
In spatial statistics, an important problem is how to represent spatial models in a way that is computationally efficient, accurate, and convenient to use. Models in R‐INLA focus on sparse precision (inverse covariance) matrices to compute inference quickly. Hence, our implementations of spatial models focus on how to represent the spatial field in such a way that the precision matrix for the "representation" is very sparse. This graphic shows a representation of a Norwegian fjord with a mesh, from which basis functions are built in the finite element method. We use sums of these basis functions to represent the spatial field. This representation has many advantages, but requires some mathematical effort to understand and to set up.
Abstract Full article on Wiley Online Library:   HTML | PDF

Focus Articles

Time series clustering and classification via frequency domain methods
Published Online: Jul 20 2018
DOI: 10.1002/wics.1444
Clustering/classification of time series Group 1 (earthquake) versus Group 2 (explosion). Data are eqexp in R package astsa.
Abstract Full article on Wiley Online Library:   HTML | PDF
Weighted cross validation in model selection
Published Online: Jul 09 2018
DOI: 10.1002/wics.1439
Classical cross‐validation method for model selection and its robustification via weighted likelihood estimating equations method.
Abstract Full article on Wiley Online Library:   HTML | PDF

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts