A1176
Title: Bayesian geostatistics using predictive stacking
Authors: Lu Zhang - University of Southern California (United States) [presenting]
Wenpin Tang - Columbia University (United States)
Sudipto Banerjee - UCLA (United States)
Abstract: Bayesian predictive stacking is presented for geostatistical models, where the primary inferential objective is to provide inference on the latent spatial random field and conduct spatial predictions at arbitrary locations. Analytically tractable posterior distributions are exploited for regression coefficients of predictors and the realizations of the spatial process conditional upon process parameters. Such inference is subsequently combined by stacking these models across the range of values of the hyper-parameters. Stacking of means and posterior densities are devised in a manner that is computationally efficient without resorting to iterative algorithms such as Markov chain Monte Carlo (MCMC) and can exploit the benefits of parallel computations. Novel theoretical insights are offered into the resulting inference within an infill asymptotic paradigm and through empirical results showing that stacked inference is comparable to full sampling-based Bayesian inference at a significantly lower computational cost.
