Multi-scale process modelling and distributed computation for spatial data
From MaRDI portal
Computational methods for problems pertaining to statistics (62-08) Directional data; spatial statistics (62H11) Monte Carlo methods (65C05) Inference from stochastic processes and prediction (62M20) Applications of statistics to environmental and related topics (62P12) Coloring of graphs and hypergraphs (05C15)
Abstract: Recent years have seen a huge development in spatial modelling and prediction methodology, driven by the increased availability of remote-sensing data and the reduced cost of distributed-processing technology. It is well known that modelling and prediction using infinite-dimensional process models is not possible with large data sets, and that both approximate models and, often, approximate-inference methods, are needed. The problem of fitting simple global spatial models to large data sets has been solved through the likes of multi-resolution approximations and nearest-neighbour techniques. Here we tackle the next challenge, that of fitting complex, nonstationary, multi-scale models to large data sets. We propose doing this through the use of superpositions of spatial processes with increasing spatial scale and increasing degrees of nonstationarity. Computation is facilitated through the use of Gaussian Markov random fields and parallel Markov chain Monte Carlo based on graph colouring. The resulting model allows for both distributed computing and distributed data. Importantly, it provides opportunities for genuine model and data scaleability and yet is still able to borrow strength across large spatial scales. We illustrate a two-scale version on a data set of sea-surface temperature containing on the order of one million observations, and compare our approach to state-of-the-art spatial modelling and prediction methods.
Recommendations
- Spatial Process Simulation
- A modeling approach for large spatial datasets
- A model for large multivariate spatial data sets
- scientific article; zbMATH DE number 3965276
- scientific article; zbMATH DE number 1346455
- scientific article; zbMATH DE number 168075
- A Data Model for Distributed Multiresolution Multisource Scientific Data
Cites work
- scientific article; zbMATH DE number 1114428 (Why is no real title available?)
- scientific article; zbMATH DE number 1134987 (Why is no real title available?)
- A Bayesian Kriged Kalman Model for Short-Term Forecasting of Air Pollution Levels
- A Full Scale Approximation of Covariance Functions for Large Spatial Data Sets
- A theoretical analysis of backtracking in the graph coloring problem
- An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach
- Bayesian computation and stochastic systems. With comments and reply.
- Blind source separation for spatial compositional data
- Computational Physics
- Data-Driven Spatio-Temporal Modeling Using the Integro-Difference Equation
- Efficient Algorithms for Bayesian Nearest Neighbor Gaussian Processes
- Fitting Gaussian Markov Random Fields to Gaussian Fields
- Fixed Rank Kriging for Very Large Spatial Data Sets
- Formal proof - the four color theorem
- Gaussian Markov Random Fields
- Gaussian Predictive Process Models for Large Spatial Data Sets
- Going off grid: computationally efficient inference for log-Gaussian Cox processes
- On Block Updating in Markov Random Field Models for Disease Mapping
- On computation using Gibbs sampling for multilevel models
- Parallel inference for massive distributed spatial data using low-rank models
- Parameter estimation in high dimensional Gaussian distributions
- Partially Collapsed Gibbs Samplers
- Posterior inference for sparse hierarchical non-stationary models
- Sampling Strategies for Fast Updating of Gaussian Markov Random Fields
- Statistics for spatio-temporal data
- Strictly Proper Scoring Rules, Prediction, and Estimation
- Variational Estimation in Spatiotemporal Systems From Continuous and Point-Process Observations
Cited in
(10)- Discussion on: ``A high-resolution bilevel skew-\(t\) stochastic generator for assessing Saudi Arabia's wind energy resources
- Fitting large-scale structured additive regression models using Krylov subspace methods
- scientific article; zbMATH DE number 1870067 (Why is no real title available?)
- Modeling and analysis of a Canadian Forces Geomatics division workflow
- Principles for statistical inference on big spatio-temporal data from climate models
- Multi-scale shotgun stochastic search for large spatial datasets
- Large multi-scale spatial modeling using tree shrinkage priors
- Multivariate spatial meta kriging
- Spatial Process Simulation
- Parallel inference for massive distributed spatial data using low-rank models
This page was built for publication: Multi-scale process modelling and distributed computation for spatial data
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2209724)
