Motivated by the problem of modelling high-dimensional multivariate referenced data, arising in many areas of research, this article proposes a Generalised Latent Spatial Quantile Regression (GLSQR) model as a reliable solution in studying the effects of some covariates across the quantiles of the response distribution. In addition, the discussed model warrants consideration when the matrix of the explanatory variables is defined through a set of spatial common latent factors. The latent factors and quantile regression components are estimated through a hierarchical Bayesian procedure and MCMC algorithms are used to provide full probabilistic inference.We illustrate the use of our GLSQR model with application to radon data, as motivating example coming from the environmental protection research.
Bayesian Structural Equation Modeling for Factors Influencing Residential Radon Levels
IPPOLITI, Luigi;FONTANELLA, Lara;VALENTINI, PASQUALE;SARRA A.;
2014-01-01
Abstract
Motivated by the problem of modelling high-dimensional multivariate referenced data, arising in many areas of research, this article proposes a Generalised Latent Spatial Quantile Regression (GLSQR) model as a reliable solution in studying the effects of some covariates across the quantiles of the response distribution. In addition, the discussed model warrants consideration when the matrix of the explanatory variables is defined through a set of spatial common latent factors. The latent factors and quantile regression components are estimated through a hierarchical Bayesian procedure and MCMC algorithms are used to provide full probabilistic inference.We illustrate the use of our GLSQR model with application to radon data, as motivating example coming from the environmental protection research.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
