Gaussian processes for machine learning in R and FORTRAN.
Gaussian Processes have recently gained a lot of attention in machine learning. gpR
shows how the calculation of the posterior predictive of a Gaussian Process and prediction of novel data is done when the kernel parameters are known. In the next versions I will implement how those are calculated by optimizing the marginal likelihood and probably include more kernels.
Install gpR
using:
devtools::install_github("dirmeier/gpR")
from the R-console.
Load the package using library(gpR)
. We provide a vignette for the package that can be called using: vignette("gpR")
. This should be all the information you need. For regression try the demo-tour using:
demo.regression()
or for classification (i.e. binomial responses):
demo.bin.classification()
Also check out the source code for more info, fork the package, or just write me!
- Simon Dirmeier [email protected]