What attribute role is "Batch"?

Best Answer

• Options

  earmijo Member Posts: 270 Unicorn

  January 2017 Solution Accepted

  When you use cross-validation RM separates the samples into K partitions at random (or in a stratified way). If you want to have full control of how the observations are assigned to the different partitions you use batch. This is going to be a column with numbers from 1 to K indicating the partition to which each observation belongs.

   

  I've used batch cross-validation when I want to compare the results from Rapidminer to the results of other programs. 

  1