Machine Learning Metaprogramming for R
These functions export several kinds of machine learning models trained in R to other programming languages, so the models can score data in a separate environment without R. Currently Base SAS is supported, and additional SAS licenses for SAS/IML, SAS/STAT, and SAS Enterprise Miner are not required. In some cases the models in SAS run faster and use less memory than in R.
Supported models:
- Bagged MARS from
caret
- Conditional inference trees from
party
- Random forests from
party
- Multivariate Adaptive Regression Splines (MARS) from
earth
- Generalized Boosted Regression Models from
gbm
- Artificial neural networks from
nnet
Later I will post this package on CRAN. Until then, use:
require(devtools)
install_github("az0/mlmeta")
For an example usage, please see:
If you find these useful, please consider a donation to Compassion International. Even a penny will help.