Bayesian Spatial Modelling with R-INLA

By Finn Lindgren, Håvard Rue. 2015

1,259 citations recorded by the source corpus.

The principles behind the interface to continuous domain spatial models in the RINLA software package for R are described. The integrated nested Laplace approximation (INLA) approach proposed by Rue, Martino, and Chopin (2009) is a computationally effective alternative to MCMC for Bayesian inference. INLA is designed for latent Gaussian models, a very wide and flexible class of models ranging from (generalized) linear mixed to spatial and spatio-temporal models. Combined with the stochastic partial differential equation approach (SPDE, Lindgren, Rue, and Lindstrom 2011), one can accommodate all kinds of geographically referenced data, including areal and geostatistical ones, as well as spatial point process data. The implementation interface covers stationary spatial models, non-stationary spatial models, and also spatio-temporal models, and is applicable in epidemiology, ecology, environmental risk assessment, as well as general geostatistics.

Subfields: Artificial Intelligence, Environmental Engineering, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/f8bece4eb99311ba5da5772e598e1985a7c3bcf4

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Environmental Engineering

In Statistics and Probability

Citations in the corpus

Estimation of a non-stationary model for annual precipitation in southern Norway using replicates of the spatial field (2014, methodology): “We refer to Ingebrigtsen et al. (2014, Section 4.5) for details on inference with our precipitation models using INLA, and to Lindgren and Rue (2014) for a thorough introduction to Bayesian spatial modelling with the SPDE approach and INLA.”
Spatial factor analysis: a new tool for estimating joint species distributions and correlations in species range (2015, methodology): “We use R-INLA (Lindgren & Rue 2015) to compute the three components of the precision matrix necessary for the stochastic partial differential equation approximation toGRFs (Lindgren, Rue & Lindstr€om 2011).”
Basis-Function Models in Spatial Statistics (2021, positive): “INLA is a general-purpose package, not specifically designed with spatial prediction in mind; a recently developed package aimed at facilitating spatial modelling and fitting with INLA is inlabru (Bachl et al., 2019).”
Fast Calibrated Additive Quantile Regression (2017, methodology): “The dataset is available within the R-INLA R package (Lindgren and Rue, 2015).”
Bayesian Computing with INLA: A Review (2016, methodology): “The dual interpretation of Matérn fields, both using covariances and also using its Markov properties, is very convenient both from a computational but also from a statistical modeling point of view (Simpson et al., 2011, 2012; Lindgren and Rue, 2015).”
Multivariate covariance generalized linear models (2015, methodology): “In the Bayesian context the flexible packages INLA (Rue et al.; 2014) and MCMCpack (Martin et al.; 2011) do not deal with multivariate response variables, judging from the package documentation.”
Spatiotemporal wildfire modeling through point processes with moderate and extreme marks (2021, methodology): “Integrated nested Laplace approximation (INLA Rue et al. 2009, Lindgren & Rue 2015, Opitz 2017) is a Bayesian technique for generalized additive models with Gaussian random effects.”
Maximum likelihood estimation of spatially varying coefficient models for large data with an application to real estate price prediction (2020, methodology): “The SPDE method cannot be considered as there are too many hyper parameters for INLA.”
INLA goes extreme: Bayesian tail regression for the estimation of high spatio-temporal quantiles (2018, background): “Each temporal effect, denoted by x(t) for simplicity, is assumed to be driven by a second-order Gaussian random walk prior x (ω) defined over weeks with precision τt > 0 for its weekly innovations (Lindgren and Rue 2008, and Rue and Held 2005, Chapter 3). Let ωt = {1, . . . , n/7 } denote the week a”
Spatial modelling with R-INLA: A review (2018, methodology): “However, if there are several other model components or a space-time model component, keeping track of all the diﬀerent A -matrices is challenging, and we developed the inla.stack interface to handle all these matrix operations in a general way (Lindgren and Rue, 2015).”