How can I effectively use the R prediction model from Java?

Question

I have some classification models that I developed in R with features such as glm, rpart, etc. What is the most computationally efficient way to name these models with Java? I saw a JRI, but it looks like there are many text calls to R.

Is there a way to use these models with Java with low overhead?

Answer 1

JPMML now has a functional testing module that specifically evaluates PMML models that were developed using R / Rattle: https://github.com/jpmml/jpmml/tree/master/pmml-rattle

JPMML should be able to clog decision trees (i.e. the rpart () function) and neural networks (i.e. the nnet () function) without any problems. Support for generic regression models (i.e. the glm () function) is coming soon.

Answer 2

Do you find it too slow to export them with the pmml package and then use a java pmsl reader like jpmml or Zementis? http://code.google.com/p/jpmml/