Create R Model with Evaluation

February 22, 2016

384 views
281 downloads


Report Abuse
This experiment demonstrates how to use the "Create R Model" module to train, score and evaluate a binary classification model
This experiment illustrates the use of the "Create R Model" to train and score a simple Naïve Bayes binary classification model. In addition, by using a few simple functions in the scoring script, score columns in the dataset can be tagged. This in turn allows the built-in evaluation module to correctly produce the appropriate evaluation metrics for the R model (in this case, the RoC curve).