validation in bootstrap forest

Question

Attached are two jmp files with same data. The only difference is how I created the validation column. In file "bootstrap test" I created validation following steps metioned in JMP documentation. Here is the link to that page

https://www.jmp.com/support/help/14-2/example-of-the-make-validation-column-utility.shtml

And in file "bootstrap test_autovalidation" I used add-in I found on JMP user community page. Here is the link

https://community.jmp.com/t5/JMP-Add-Ins/Add-in-To-Support-Auto-Validation-Workflow/ta-p/189991

But the results I got from both the analysis is different. Is this difference because of the validation column was setup or something else? Example of difference in results such as if I look at column contribution both the methods give me different results.

bootstrap test.jmp

bootstrap test_Autovalidation.jmp

Mark_Bailey · Accepted Answer

The assignment of observations in the sample to the training or validation sets is random. So you would not expect the same results.

Mark_Bailey · Accepted Answer

The first example uses 50% hold out for validation. The second example uses 25% hold out for validation. I would not expect to get the same results.