Altair AI studio: feeding a validation set into the split validation operator
Description: I would like to use the split validation operator in Altair AI Studio, but with a caveat: my validation set is already separate from the train set (that is, I load them from separate csv files). The description of the split validation operator suggests that I should be able to do this ("The Split Validation operator also allows training on one data set and testing on another explicit testing data set." - https://docs.rapidminer.com/latest/studio/operators/validation/split_validation.html ) the problem is that there is no further specification on how exactly this should be done. One would expect that there would be an optional "test" input to the operator, but this is not the case. But then how would one feed the explicit testing data set into the operator? Thank you in advance for any help.
Product/Topic Name : Altair AI Studio
Hi György,
As you are probably aware, for the Split Validation operator, you feed it a single data set
Then within the operator it feeds one side a randomly selected Training subset, and feeds the remainder of the data set to the Testing side
You would like to specify your own training and testing data set. Let's say that these data sets are called Titanic Training and Titanic Unlabeled (the latter is your testing data). To use those data sets instead of the ones passed into the operator, you can just retrieve them from within either side of the operator, as shown here:
Although it is not clear why you would do this since you are not using the operator for its purpose and since you could do the same thing without the operator.