Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Deprecate Scikit-Learn based Ensembling Component #2620

Closed
christopherbunn opened this issue Aug 12, 2021 · 0 comments · Fixed by #2819
Closed

Deprecate Scikit-Learn based Ensembling Component #2620

christopherbunn opened this issue Aug 12, 2021 · 0 comments · Fixed by #2819
Labels
enhancement An improvement to an existing feature.

Comments

@christopherbunn
Copy link
Contributor

christopherbunn commented Aug 12, 2021

As part of our implementation of our new ensembler component, we will need to deprecate the current scikit-learn based StackingEnsembler. Once we finish the new ensembler component, it will be used alongside the scikit-learn version during an ensembling AutoML search. However, because the new ensembler component's will contain the same functionality of the scikit-learn based ensembler, there isn't a need to keep both. This issue tracks deprecating the scikit-learn based ensembler.

Current implementation plan:

  • Keep both ensemblers in AutoML search, but raise a DeprecationWarning during search that the sklearn version will not be included in search during the next release.
  • Remove the scikit-learn based ensembler from AutoML search, but still retain
    • This will allow users to create new pipelines with this component manually if necessary
  • Eventually, remove the scikit-learn based ensembler component from EvalML completely
@christopherbunn christopherbunn added the enhancement An improvement to an existing feature. label Aug 12, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement An improvement to an existing feature.
Projects
None yet
Development

Successfully merging a pull request may close this issue.

1 participant