FEA Add positive and negative likelihood ratios to metrics #22518
Conversation
This file contains 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
ArturoAmorQ
changed the title
jjerphan
changed the title
jjerphan
reviewed
Mar 4, 2022
jjerphan
reviewed
Mar 10, 2022
…rn into likelihood_ratios
Co-authored-by: Julien Jerphanion <git@jjerphan.xyz>
GaelVaroquaux
reviewed
Mar 30, 2022
GaelVaroquaux
reviewed
Mar 30, 2022
GaelVaroquaux
reviewed
Mar 30, 2022
The example shows how the decision boundary is kept constant across different prevalences. I thought the illustration was required as this is indeed not the most intuitive user case.
Good idea, it's done :) |
Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
GaelVaroquaux
approved these changes
May 10, 2022
GaelVaroquaux
reviewed
May 10, 2022
…nto likelihood_ratios
Co-authored-by: Gael Varoquaux <gael.varoquaux@normalesup.org>
…rn into likelihood_ratios
Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
|
It seems that you still have a conflict :( |
glemaitre
reviewed
May 13, 2022
Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
…nto likelihood_ratios
…rn into likelihood_ratios
glemaitre
approved these changes
May 17, 2022
|
Thanks @jjerphan, @GaelVaroquaux, @glemaitre and @ogrisel for your time and the fruitful discussions we had online and on real life. It was really fun to implement this feature! |
|
A great addition, and great docs to push users to think about what they
do.
|
lesteve
pushed a commit
to lesteve/scikit-learn
that referenced
this pull request
May 19, 2022
…arn#22518) Co-authored-by: Julien Jerphanion <git@jjerphan.xyz> Co-authored-by: Olivier Grisel <olivier.grisel@ensta.org> Co-authored-by: Gael Varoquaux <gael.varoquaux@normalesup.org> Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
mathijs02
pushed a commit
to mathijs02/scikit-learn
that referenced
this pull request
Dec 27, 2022
…arn#22518) Co-authored-by: Julien Jerphanion <git@jjerphan.xyz> Co-authored-by: Olivier Grisel <olivier.grisel@ensta.org> Co-authored-by: Gael Varoquaux <gael.varoquaux@normalesup.org> Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
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.
Reference Issues/PRs
Fixes #22404.
What does this implement/fix? Explain your changes.
We agreed that we should add positive and negative likelihood ratios as they are considered as standard good practice in medicine / public health, as well as interpretable in terms of the pre-test versus post-test odds ratio even in class imbalance.
Any other comments?
This is a first PR that creates a broad function that computes both positive and negative likelihood ratios. A second step of creating functions that specifically select one or the other ("positive_likelihood_ratio" and a "negative_likelihood_ratio") will be addressed in another PR once that we agreed on forms and variable names for this one.