deepgp: Deep Gaussian Processes using MCMC

Performs posterior inference for deep Gaussian processes following Sauer, Gramacy, and Higdon (2020) <arXiv:2012.08015>. Models are trained through MCMC including elliptical slice sampling of latent Gaussian layers and Metropolis-Hastings sampling of kernel hyperparameters. Vecchia-approximation for faster computation is implemented following Sauer, Cooper, and Gramacy (2022) <arXiv:2204.02904>. Downstream tasks include sequential design through active learning Cohn/integrated mean squared error (ALC/IMSE; Sauer, Gramacy, and Higdon, 2020) and optimization through expected improvement (EI; Gramacy, Sauer, and Wycoff, 2021 <arXiv:2112.07457>). Models extend up to three layers deep; a one layer model is equivalent to typical Gaussian process regression. Covariance kernel options are matern (default) and squared exponential. Applicable to both noisy and deterministic functions. Incorporates SNOW parallelization and utilizes C and C++ under the hood.

Version: 1.0.0
Depends: R (≥ 3.6)
Imports: grDevices, graphics, stats, doParallel, foreach, parallel, GpGp, Matrix, Rcpp, mvtnorm, FNN
LinkingTo: Rcpp, RcppArmadillo
Suggests: akima, knitr
Published: 2022-04-08
Author: Annie Sauer
Maintainer: Annie Sauer <anniees at vt.edu>
License: LGPL-2 | LGPL-2.1 | LGPL-3 [expanded from: LGPL]
NeedsCompilation: yes
Materials: README
CRAN checks: deepgp results

Documentation:

Reference manual: deepgp.pdf

Downloads:

Package source: deepgp_1.0.0.tar.gz
Windows binaries: r-devel: deepgp_1.0.0.zip, r-release: deepgp_1.0.0.zip, r-oldrel: deepgp_1.0.0.zip
macOS binaries: r-release (arm64): deepgp_1.0.0.tgz, r-oldrel (arm64): deepgp_1.0.0.tgz, r-release (x86_64): deepgp_1.0.0.tgz, r-oldrel (x86_64): deepgp_1.0.0.tgz
Old sources: deepgp archive

Linking:

Please use the canonical form https://CRAN.R-project.org/package=deepgp to link to this page.