cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Choose Language Hide Translation Bar
0 Kudos

K-fold crossvalidation should be an option on all platforms

Right now I can do K-fold crossvalidation when building a neural network:

dmmdiego_0-1642105633361.png

 

And with the recent XGBoost add-in (https://community.jmp.com/t5/JMP-Add-Ins/XGBoost-Add-In-for-JMP-Pro/ta-p/319383), it's even better as multiple folds can be added, which are orthogonal to each other:

dmmdiego_1-1642105693808.png

 

dmmdiego_2-1642105720697.png

However, this option is not available for other platforms, such as Boosted Tree, Boostrap Forest, or even a simpler Multiple Linear Regression model.

Only you can add a single validation column with a unique hold-out sample.

 

There is nothing unique about k-fold crossvalidation to XGBoost, so this option should be available for all platforms, to allow a more generalized error metrics across models consistently evaluated across multiple platforms.