Fix: Use float64 epsilon when clipping initial probabilities in GradientBoosting #31575
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…entBoosting Replaced `np.finfo(np.float32).eps` with `np.finfo(np.float64).eps` in `sklearn/ensemble/_gb.py::_get_init_raw_predictions`. Why: * Predictions are promoted to `float64` immediately after clipping; using a `float32` epsilon (≈1e-7) was unnecessarily large. * Large epsilon shrank extreme probabilities (near 0 or 1), degrading log-loss and calibration, especially on well-separated classes. Impact: * Higher numerical fidelity of initial raw predictions. * Better probability calibration and potential convergence improvements. * Removes obsolete “FIXME” comment.
|
I wonder why it was |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Replaced
np.finfo(np.float32).epswithnp.finfo(np.float64).epsinsklearn/ensemble/_gb.py::_get_init_raw_predictions.Why:
float64immediately after clipping; using afloat32epsilon (≈1e-7) was unnecessarily large.Impact: