A Divide-and-Conquer Bayesian Approach to Large-Scale Kriging

Abstract

We propose a three-step divide-and-conquer strategy within the Bayesian paradigm that delivers massive scalability for any spatial process model. We partition the data into a large number of subsets, apply a readily available Bayesian spatial process model on every subset, in parallel, and optimally combine the posterior distributions estimated across all the subsets into a pseudo-posterior distribution that conditions on the entire data. The combined pseudo posterior distribution replaces the full data posterior distribution for predicting the responses at arbitrary locations and for inference on the model parameters and spatial surface. Based on distributed Bayesian inference, our approach is called "Distributed Kriging" (DISK) and offers significant advantages in massive data applications where the full data are stored across multiple machines. We show theoretically that the Bayes $L_2$-risk of the DISK posterior distribution achieves the near optimal convergence rate in estimating the true spatial surface with various types of covariance functions, and provide upper bounds for the number of subsets as a function of the full sample size. The model-free feature of DISK is demonstrated by scaling posterior computations in spatial process models with a stationary full-rank and a nonstationary low-rank Gaussian process (GP) prior. A variety of simulations and a geostatistical analysis of the Pacific Ocean sea surface temperature data validate our theoretical results.

arXiv comment29 pages, including 4 figures and 5 tables

References in corpus (9)

• Fast Computation of Wasserstein Barycenters
• Rates of contraction of posterior distributions based on Gaussian process priors
• Divide and Conquer Kernel Ridge Regression: A Distributed Algorithm with Minimax Optimal Rates
• Adaptive Bayesian estimation using a Gaussian random field with inverse Gamma bandwidth
• Robust and Scalable Bayes via a Median of Subset Posterior Measures
• Frequentist coverage and sup-norm convergence rate in Gaussian process regression
• An asymptotic analysis of distributed nonparametric methods
• Parallel Streaming Wasserstein Barycenters
• A Case Study Competition Among Methods for Analyzing Large Spatial Data

Cited by in corpus (7)

• A Case Study Competition Among Methods for Analyzing Large Spatial Data
• Distributed Bayesian Varying Coefficient Modeling Using a Gaussian Process Prior
• Deep Gaussian Markov Random Fields
• Divide-and-conquer methods for big data analysis
• A Divide and Conquer Algorithm of Bayesian Density Estimation
• A Fused Gaussian Process Model for Very Large Spatial Data
• An Algorithm for Distributed Bayesian Inference in Generalized Linear Models