User 1129 | 12/30/2014, 8:36:04 AM
is there a reason that gl.boostedtreesclassifier.create accepts argument "validationset", while gl.logisticclassifier.create does not?
User 940 | 12/30/2014, 9:37:15 PM
Thank you for pointing this out. We are always trying to make our interface more consistent, and we'll work on fixing this.
You can always evaluate the already-trained model on the validation set with .evaluate(). Would this be a workaround that is OK for the time being?
User 1129 | 12/31/2014, 8:59:25 AM
yes, I'm aware of this workaround. Thank you