What inspired this Wish List Request?
Yesterday, I attended a seminar about desicion trees and ensample learning for beginners held by The Institute of Statistical Mathematics in Japan. The instructor is Prof. Icihgaku Takigawa. He used only Python in his seminar.
There are three methods which I am intereted in, but aren't implmented in JMP Pro.
I entered the third one of them here.
Please describe the new or improved feature or solution you would like to see.
An option to calculate prediction precision (aka Uncertainty Quantification) for Random Forest or Boosted Trees.
What would the value be to you and/or other users if this new feature is provided?
Sometimes, we want to see prediction precision as well as as point prediction. This feauture can help users to know how the prediction variability is large.
Is this an external customer request?
Is there a competitive software package that has this feature or solution that we should be aware of?
The calculation of standard errors for predictions in Random Forest is easy. Please se _return_std function in the following page.
https://scikit-optimize.github.io/stable/_modules/skopt/learning/forest.html
It's very complicated to calculate standard errors of predictions in Boosted Trees. Prof. Takigawa introduced MAPIE, but I have not followed the information. The URL of the homepage of MAPIE is the following.
https://mapie.readthedocs.io/en/latest/index.html
Will this impact a renewal, expansion, or adoption at a customer, and does this put an account at risk if we do not address this?
This is not requested by users.
