I'm working with a dataset of about 27,000 rows and am trying to use 9 factors to predict a response using the partition function and a decision tree. I'm using both k-means and validation. When I tell JMP to Go, it will split 100-200 times, however the Split History shows me very little improvement after about 20 splits.
For what I'm trying to do, I want to back up to the 20th split, but I don't see a way to do that without either starting fresh and manually splitting 20 times, or starting at the end and manually pruning 180 times. Is there a way to select the conditions on a specific split number?