Roman Lutz

I'm open to other names. `Thresholder` comes to mind, but I'm sure there are better names. > > Given the similarities I think the entire API should basically be identical...

The more I think about it I feel like we should do the generic one I mentioned, and then have the Reject Option Classifier additionally as a special case of...

I agree with everything you wrote @hildeweerts. I think @MiroDudik just meant that you don't have an algorithm that computes the thresholds for you, so you may end up trying...

Hi @mnalk ! To use code from the version you can install from PyPI you'll have to wait a little while because #974 is not merged yet, which means it...

@Hadyark do you mean for Fairlearn? @hildeweerts tagged it as help wanted so you should feel free to draft a PR. I would recommend publishing it early on to get...

No, we don't have a custom implementation. What you wrote matches my expectations, just wanted to clarify to make sure we are thinking of the same thing. 🙂

I don't see a problem with that, but I also haven't worked with the internals of sklearn's decision trees before. Tagging @fairlearn/fairlearn-maintainers to chime in if they have thoughts.

Definitely seems like an openml issue to me, and not because of removed datasets but unavailable service. PapersWithCode also recently added datasets: https://paperswithcode.com/datasets

Thanks for linking @hildeweerts . Sounds like this should be resolved now. Can we close this issue?

FWIW I haven't seen this in quite a while. Thanks @hildeweerts for raising it with OpenML folks, and obviously thanks to the OpenML folks who evidently made a few improvements...