[MRG] DOC add permutation importance to GradientBoostingRegressor#16742
[MRG] DOC add permutation importance to GradientBoostingRegressor#16742adrinjalali merged 5 commits intoscikit-learn:masterfrom
Conversation
200cbd4 to
5d89ab6
Compare
5d89ab6 to
7b02f05
Compare
There was a problem hiding this comment.
LGTM, can you just please add a small note, either in the top level comment or an inline comment just before or right after the plot to give a short analysis of the results: in particular the two methods agree to identify the same top 2 features as strongly predictive features but not in the same order.
The third most predictive feature, "bp", is also the same for the 2 methods. The remaining features are less predictive and the error bars of the permutation plot show that they overlap with 0.
266cf89 to
117ad5d
Compare
117ad5d to
b478aee
Compare
|
@ogrisel Thanks for the suggestion! Note added. |
adrinjalali
left a comment
There was a problem hiding this comment.
Thanks @nilichen. Other than the minor point LGTM.
Please also refrain from force pushing an amended history instead of pushing new commits. We'll squash and merge and all your commits will become one at the end anyway :)
Co-Authored-By: Adrin Jalali <adrin.jalali@gmail.com>
|
@adrinjalali Got it. |
…earn#16742) * add permutation importance to gbr * format document * separate deviance from feature importance during plotting * add some analysis * Update examples/ensemble/plot_gradient_boosting_regression.py Co-Authored-By: Adrin Jalali <adrin.jalali@gmail.com> Co-authored-by: Katrina Ni <kani@tableau.com> Co-authored-by: Adrin Jalali <adrin.jalali@gmail.com>
Reference Issues/PRs
Towards #14528.
What does this implement/fix? Explain your changes.
The latest doc: https://97558-843222-gh.circle-artifacts.com/0/doc/auto_examples/ensemble/plot_gradient_boosting_regression.html
The previous way of having one plot but separate code blocks didn't really render as expected (https://97540-843222-gh.circle-artifacts.com/0/doc/auto_examples/ensemble/plot_gradient_boosting_regression.html)