Evaluation of R multi-class classification models

August 19, 2016

450 views
159 downloads


Report Abuse
Demonstrates the use of "Create R Model" to train and score a multi-class classifier.
This experiment demonstrates the use of the `set.multiclass.classification.scores` function in AzureML to set the schema of scores returned by a multi-class classifier trained in R. The scored dataset with true labels can then be passed to Evaluate Model to produce the confusion matrix and other multi-class evaluation metrics