<b>tgp</b>: An<i>R</i>Package for Bayesian Nonstationary, Semiparametric Nonlinear Regression and Design by Treed Gaussian Process Models

Gramacy, Robert B.

doi:10.18637/jss.v019.i09

Cited by 172 publications

(130 citation statements)

References 33 publications

Supporting

Mentioning

130

Contrasting

Order By: Relevance

“…We keep it smaller here for faster execution in live demonstration. Observe that the code uses uses lhs from the tgp package (Gramacy 2007;Gramacy and Taddy 2010), rather than from lhs, because the tgp version allows a non-unit rectangle, which is required for our second use of lhs below.…”

Section: An Illustrative Examplementioning

confidence: 99%

“…There are several packages on the Comprehensive R Archive Network (CRAN, https://CRAN.R-project.org/) which implement full (i.e., not approximated) Gaussian process regression. These include mleGP (Dancik 2013), GPfit (MacDonald, Ranjan, and Chipman 2015), spatial (Venables and Rip-ley 2002), and fields (Nychka, Furrer, and Sain 2016) -all performing maximum likelihood (or a posteriori) inference; and tgp (Gramacy 2007;Gramacy and Taddy 2010) and spBayes (Finley, Banerjee, and Carlin 2007;Finley, Banerjee, and E.Gelfand 2015) -performing fully Bayesian inference. Approximate methods for large-scale inference include tgp and sparseEM (Kaufman, Bingham, Habib, Heitmann, and Frieman 2012, which is not on CRAN).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

laGP: Large-Scale Spatial Modeling via Local Approximate Gaussian Processes inR

Gramacy¹

2016

J. Stat. Soft.

Self Cite

155

142

View full text Add to dashboard Cite

Gaussian process (GP) regression models make for powerful predictors in out of sample exercises, but cubic runtimes for dense matrix decompositions severely limit the size of data -training and testing -on which they can be deployed. That means that in computer experiment, spatial/geo-physical, and machine learning contexts, GPs no longer enjoy privileged status as data sets continue to balloon in size. We discuss an implementation of local approximate Gaussian process models, in the laGP package for R, that offers a particular sparse-matrix remedy uniquely positioned to leverage modern parallel computing architectures. The laGP approach can be seen as an update on the spatial statistical method of local kriging neighborhoods. We briefly review the method, and provide extensive illustrations of the features in the package through worked-code examples. The appendix covers custom building options for symmetric multi-processor and graphical processing units, and built-in wrapper routines that automate distribution over a simple network of workstations.

show abstract

Section: An Illustrative Examplementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

laGP: Large-Scale Spatial Modeling via Local Approximate Gaussian Processes inR

Gramacy¹

2016

J. Stat. Soft.

Self Cite

155

142

View full text Add to dashboard Cite

show abstract

“…The tgp package (Gramacy 2007;Gramacy and Lee 2008), originally developed for building surrogates of both stationary and non-stationary noisy simulators, uses a GP model for emulating the stationary components of the process. The GP model here includes a nugget parameter that is estimated along with other parameters.…”

Section: Comparison With Other Packagesmentioning

confidence: 99%

GPfit: AnRPackage for Fitting a Gaussian Process Model to Deterministic Simulator Outputs

MacDonald¹,

Ranjan²,

Chipman³

2015

J. Stat. Soft.

View full text Add to dashboard Cite

Gaussian process (GP) models are commonly used statistical metamodels for emulating expensive computer simulators. Fitting a GP model can be numerically unstable if any pair of design points in the input space are close together. Ranjan, Haynes, and Karsten (2011) proposed a computationally stable approach for fitting GP models to deterministic computer simulators. They used a genetic algorithm based approach that is robust but computationally intensive for maximizing the likelihood. This paper implements a slightly modified version of the model proposed by Ranjan et al. (2011) in the R package GPfit. A novel parameterization of the spatial correlation function and a clustering based multistart gradient based optimization algorithm yield robust optimization that is typically faster than the genetic algorithm based approach. We present two examples with R codes to illustrate the usage of the main functions in GPfit. Several test functions are used for performance comparison with the popular R package mlegp. We also use GPfit for a real application, i.e., for emulating the tidal kinetic energy model for the Bay

show abstract

“…Hyperparameters µ, B, V, ρ, α σ , (Gramacy and Lee, 2008). We use the tgp package in R for implementing the Treed Gaussian Process and related analyses (Gramacy, 2007). 7…”

Section: Treed Gaussian Processmentioning

confidence: 99%

Variable selection via a multi-stage strategy

Chang

Lee

2014

Journal of Applied Statistics

View full text Add to dashboard Cite

Variable selection for nonlinear regression is a complex problem, made even more difficult when there are a large number of potential covariates and a limited number of datapoints. We propose herein a multi-stage method that combines state of the art techniques at each stage to best discover the relevant variables. At the first stage, an extension of the Bayesian Additive Regression tree is adopted to reduce the total number of variables to around 30. At the second stage, sensitivity analysis in Treed Gaussian Process is adopted to further reduce the total number of variables. Two stopping rules are designed and sequential design is adopted to make best use of previous information. We demonstrate our approach on two simulated examples and one real dataset.

show abstract

tgp: AnRPackage for Bayesian Nonstationary, Semiparametric Nonlinear Regression and Design by Treed Gaussian Process Models

Cited by 172 publications

References 33 publications

laGP: Large-Scale Spatial Modeling via Local Approximate Gaussian Processes inR

laGP: Large-Scale Spatial Modeling via Local Approximate Gaussian Processes inR

GPfit: AnRPackage for Fitting a Gaussian Process Model to Deterministic Simulator Outputs

Variable selection via a multi-stage strategy

Contact Info

Product

Resources

About