Massively parallel approximate Gaussian process regression
From MaRDI portal
Abstract: We explore how the big-three computing paradigms -- symmetric multi-processor (SMC), graphical processing units (GPUs), and cluster computing -- can together be brought to bare on large-data Gaussian processes (GP) regression problems via a careful implementation of a newly developed local approximation scheme. Our methodological contribution focuses primarily on GPU computation, as this requires the most care and also provides the largest performance boost. However, in our empirical work we study the relative merits of all three paradigms to determine how best to combine them. The paper concludes with two case studies. One is a real data fluid-dynamics computer experiment which benefits from the local nature of our approximation; the second is a synthetic data example designed to find the largest design for which (accurate) GP emulation can performed on a commensurate predictive set under an hour.
Recommendations
- scientific article; zbMATH DE number 6276239
- Parallel cross-validation: a scalable fitting method for Gaussian process models
- Approximate inference for spatial functional data on massively parallel processors
- Scaled Vecchia approximation for fast computer-model emulation
- Locally induced Gaussian processes for large-scale simulation experiments
Cited in
(29)- Multiresolution functional ANOVA for large-scale, many-input computer experiments
- Approximate inference for spatial functional data on massively parallel processors
- Locally induced Gaussian processes for large-scale simulation experiments
- Large-scale local surrogate modeling of stochastic simulation experiments
- Nested aggregation of experts using inducing points for approximated Gaussian process regression
- A case study competition among methods for analyzing large spatial data
- Parametric Gaussian process regression for big data
- Parallel partial Gaussian process emulation for computer models with massive output
- Polar Gaussian processes and experimental designs in circular domains
- Computationally efficient algorithm for Gaussian process regression in case of structured samples
- Graphics card as a cheap supercomputer.
- Patchwork kriging for large-scale Gaussian process regression
- Gaussian processes for computer experiments
- Gaussian process regression using partition resampling model
- A sequential split-and-conquer approach for the analysis of big dependent data in computer experiments
- Scaled Vecchia approximation for fast computer-model emulation
- scientific article; zbMATH DE number 7164769 (Why is no real title available?)
- Emulating satellite drag from large simulation experiments
- Scalable inference for space-time Gaussian Cox processes
- GPLP: a local and parallel computation toolbox for Gaussian process regression
- scientific article; zbMATH DE number 6276239 (Why is no real title available?)
- Discussion of ``Saving storage in climate ensembles: a model-based stochastic approach
- Separable spatio-temporal kriging for fast virtual sensing
- Implementation and analysis of GPU algorithms for Vecchia approximation
- Precision aggregated local models
- High-performance statistical computing (HPSC): challenges, opportunities, and future directions
- Active Learning of Piecewise Gaussian Process Surrogates
- Parallel cross-validation: a scalable fitting method for Gaussian process models
- Calibrating a large computer experiment simulating radiative shock hydrodynamics
This page was built for publication: Massively parallel approximate Gaussian process regression
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2938459)
