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!
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!