DOC Add link to Early Stopping example in Gradient Boosting#27025
DOC Add link to Early Stopping example in Gradient Boosting#27025adrinjalali merged 10 commits intoscikit-learn:mainfrom
Conversation
adrinjalali
left a comment
There was a problem hiding this comment.
You can also add the link to the parameters in the API docs, i.e. the docstring of the HistGradientBoosting classes.
@adrinjalali, ready to review. |
lorentzenchr
left a comment
There was a problem hiding this comment.
Now we need to improve the example on early stopping 😄
I'm also thinking something similar |
|
@greyisbetter Feel free to open PRs improving this or other examples. |
adrinjalali
left a comment
There was a problem hiding this comment.
I noticed that the example only uses GradientBoostingClassifier, so this example can only be linked from GradientBoostingClassifier and GradientBoostingRegressor
|
@adrinjalali, Could you please review the recent changes when you have a moment? |
| If early stopping is False, then the training stops when the training | ||
| loss does not improve by more than tol for n_iter_no_change consecutive | ||
| passes over the training set. | ||
| Only effective when solver='sgd' or 'adam'. |
There was a problem hiding this comment.
non-related change to the PR, please revert.
There was a problem hiding this comment.
Thank you for pointing that out. I've reverted the non-related changes as per your request. Your guidance is greatly appreciated
…earn#27025) Co-authored-by: Adrin Jalali <adrin.jalali@gmail.com>
Towards #26927
This adds a reference to example of choosing fewer
n_estimatorsinGradientBoostingClassifiermodel using Early Stopping technique.