You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have tried to use Anchor to explain a RandomForestRegressor, it works with no bug, however it took longer than a RandomForestClassifier and the length of rules returned is longer expecially when the precision is low. I would like to know if Anchor is indeed expected to work for regression tasks, and if so, what values expected for the parameter 'class_names' ?
Thank you in advance!
The text was updated successfully, but these errors were encountered:
I have the same issue as the question here, so long rules returned even with a low precision for regression model. Has anybody got the answer? It seems the explainer is asking for the exact CTR prediciton...
Dear Marco,
Thank you for publishing the code!
I have tried to use Anchor to explain a RandomForestRegressor, it works with no bug, however it took longer than a RandomForestClassifier and the length of rules returned is longer expecially when the precision is low. I would like to know if Anchor is indeed expected to work for regression tasks, and if so, what values expected for the parameter 'class_names' ?
Thank you in advance!
The text was updated successfully, but these errors were encountered: