How do I create a decision tree (with leaves and branches) from Gradient Boosting results?

[gcarterIT]

Hi,

I'm using the Gradient Boosting Node to compare the results to those of the Decision Tree Node.

How can I use the results of the Gradient Boosting Node to create a decision tree (with branches and leaves) similar to that produced by the Decision Tree Node.

Thanks.

[gcarterIT]

I'm using the HP Forest Node to compare the results to those of the Decision Tree Node.

How can I use the results of the HP Forest Node to create a decision tree (with branches and leaves) similar to that produced by the Decision Tree Node.

Thanks.

[CraigDeVault · Accepted solution]

There is not a way to have the HP Forest output create a representative tree. However, you can use the Variable Importance table to see what are the variables that are most used when creating all of the trees.

[gcarterIT]

Thanks. I used the same approach for a gradient boosting model.