interpretml
sample_weight in Explainable Boosting Machine (EBM)
I am not seeing sample_weight available in Explainable Boosting Machine, is there any easy way to implement this? any direction would be really appreciated, thanks a lot!
Thanks @RealLucasMeyer! Yes, sample_weight support for EBM is also being tracked in #115 .
(Reposting from issue #62): The latest release of interpret (0.2.5) now has support for sample weights in ExplainableBoostingMachines.
You can pass in positive floating point weights to the new sample_weight parameter of the ebm.fit() call. sample_weight should be the exact same shape and dimension as y -- one weight per sample. Here's a quick usage example:
from interpret.glassbox import ExplainableBoostingRegressor
ebm = ExplainableBoostingRegressor()
ebm.fit(X, y, sample_weight=w)
You can also see more in our documentation: https://interpret.ml/docs/ebm.html#explainableboostingclassifier
To upgrade interpret using pip: pip install -U interpret
Let us know if you run into any issues! -InterpretML Team
This is great! Thanks @interpret-ml !
From my experiments on heavily imbalanced data, BalancedRandomForestClassifier with its balanced_subsample mode does perform better than EBM. Is there any way possible to compute weights based on the bootstrap sample for every tree grown similar to balanced_subsample mode?
Closing this issue as the sample_weight aspect has been resolved.
Regarding making balanced subsamples, I believe you can already achieve this using the existing interface. We use stratified sampling for classification, so every bag should have the same number of samples in each class. You can set the per-class weight through the sample_weight parameter by setting the sample weights according to their label. If there's something I'm missing that deserves a longer discussion, please open a new issue for separate tracking.
