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

Exporting EBM as PMML #536

Closed
lukedex opened this issue May 2, 2024 · 3 comments
Closed

Exporting EBM as PMML #536

lukedex opened this issue May 2, 2024 · 3 comments

Comments

@lukedex
Copy link

lukedex commented May 2, 2024

Hi all,

Has anyone had success or know of a way to export an EBM as PMML?

I can see external support to output as ONNX or SQL format but I'm struggling to find a way to export as PMML. I've looked at some of the sklearn exporting options as I thought this way be a way forward but have failed on those.

Thanks

@paulbkoch
Copy link
Collaborator

Hi @lukedex -- I'm not aware of a PMML exporter for EBMs and a search on github for "explainable boosting" PMML yields no results, so if there is one it's very well hidden.

Given the development needs in other areas I probably won't work on this myself in the near term, but I'll leave it in our backlog, and would welcome a PR if anyone would like to contribute such an exporter.

@vruusmann
Copy link

SkLearn2PMML versions 0.108.0 and newer support all InterpretML glassbox estimators, including the ExplainableBoostingClassifier and ExplainableBoostingRegressor estimator classes.

Here's a demo:

from interpret.glassbox import ExplainableBoostingClassifier
from sklearn2pmml import sklearn2pmml

classifier = ExplainableBoostingClassifier(random_state = 13)
classifier.fit(X, y)

sklearn2pmml(classifier, "EBM.pmml")

So, I believe this issue can be closed as "fixed" now.

@paulbkoch
Copy link
Collaborator

Thank you @vruusmann, that’s great news!

Closing the issue.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Development

No branches or pull requests

3 participants