evalml.objectives.FraudCost.decision_function

FraudCost.decision_function(ypred_proba, threshold=0.0, X=None)[source]

Determine if a transaction is fraud given predicted probabilities, threshold, and dataframe with transaction amount.

Parameters

  • ypred_proba (ww.DataColumn, pd.Series) – Predicted probablities

  • threshold (float) – Dollar threshold to determine if transaction is fraud

  • X (ww.DataTable, pd.DataFrame) – Data containing transaction amounts

Returns

pd.Series of predicted fraud labels using X and threshold

Return type

pd.Series

evalml.objectives.FraudCost.calculate_percent_difference evalml.objectives.FraudCost.is_defined_for_problem_type