0mlegp - Maximum Likelihood Estimation of Gaussian Processes
The package mlegp
provides maximum likelihood Gaussian process modeling for univariate and multi-dimensional outputs with diagnostic plots and sensitivity analysis.
Gaussian processes (GPs) are commonly used as surrogate statistical models for predicting output of computer experiments (Santner et al., 2003). Even more generlly, GPs are both interpolators and smoothers of data and are effective when the response surface of interest is a smooth function of the parameter space. The package finds maximum likelihood estimates of Gaussian processes for univariate and multi-dimensional responses, for Gaussian processes with product exponential correlation structures; constant or linear regression mean functions; no nugget term, constant nugget terms, or a nugget matrix that can be specifed up to a multiplicative constant. The latter is an extension of previous Gaussian process models and provides some exibility for using GPs to model heteroscedastic responses. Diagnostic plotting functions, and the sensitivity analysis tools of Functional Analysis of Variance (FANOVA) decomposition, and plotting of main and two-way factor interaction effects are implemented. Multi-dimensional output can be modelled by fitting independent GPs to each dimension of output, or to the most principle component weights following singular value decomposition of the output. Plotting of main effects for functional output is also implemented.
Once the library is installed, the library is loaded into R by calling 'library(mlegp)' from within R. A complete list of functions and vignettes can be obtained by calling `library(help = "mlegp")'.
Download mlegp from CRAN: http://cran.r-project.org/src/contrib/Descriptions/mlegp.html
Additional Information: http://www.public.iastate.edu/~gdancik/mlegp/