Lennart Purucker
Lennart Purucker
Hi @eddiebergman, Thank you very much for adding me! This is indeed a very interesting topic for ensemble building based on probabilities. --- Hi @ggjj11, We can integrate threshold-tuning into...
@jonaslandsgesell thank you for the detailed comment and input on this discussion. Your approach is very reasonable, and it is precisely what I was asking about! Originally, I thought about...
Hey, Thank you for your quick responses! @eddiebergman, I think your intuition is correct in regard to the benefits of step 4 on average. Furthermore, employing any kind of early...
Hi @mfeurer I am always happy to talk about ensembles! > Do you have any timeline on this? Not really. Depending on the progress with my current work, a few...
Hey, @mfeurer @eddiebergman I have opened a PR for use best, tie-breaking, and rounding scores (losses). I have not added anything related to a time limit, because I did not...
This was resolved by PR #1260 by adding an option to refresh the cache to `get_dataset`.
We set not downloading to be the default from 0.15.0 onwards in PR #1260.
This issue is resolved. We introduced lazy loading in #1260 and made downloading the features and qualities optional from 0.15.0 onwards.
For documentation purposes, I found the following classes that are currently missing documentation as far as I can tell: - OpenMLRun (see PR #1194) - OpenMLClusteringTask - OpenMLTask - OpenMLLearningCurveTask...
I can reproduce this issue on my cluster. Downgrading to `2.6.3` works as well (thanks for mentioning that!).