- class fairlearn.reductions.BoundedGroupLoss(loss, *, upper_bound=None)[source]#
Moment for constraining the worst-case loss by a group.
For more information refer to the user guide.
Return the number of samples in the data.
Return the vector of bounds.
Return a default objective.
Calculate the degree to which constraints are currently violated by the predictor.
load_data(X, y, *, sensitive_features)
Load data into the moment object.
Return the lambda values.
Return the signed weights.