fraud_cost ====================================== .. py:module:: evalml.objectives.fraud_cost .. autoapi-nested-parse:: Score the percentage of money lost of the total transaction amount process due to fraud. Module Contents --------------- Classes Summary ~~~~~~~~~~~~~~~ .. autoapisummary:: evalml.objectives.fraud_cost.FraudCost Contents ~~~~~~~~~~~~~~~~~~~ .. py:class:: FraudCost(retry_percentage=0.5, interchange_fee=0.02, fraud_payout_percentage=1.0, amount_col='amount') Score the percentage of money lost of the total transaction amount process due to fraud. :param retry_percentage: What percentage of customers that will retry a transaction if it is declined. Between 0 and 1. Defaults to 0.5. :type retry_percentage: float :param interchange_fee: How much of each successful transaction you pay. Between 0 and 1. Defaults to 0.02. :type interchange_fee: float :param fraud_payout_percentage: Percentage of fraud you will not be able to collect. Between 0 and 1. Defaults to 1.0. :type fraud_payout_percentage: float :param amount_col: Name of column in data that contains the amount. Defaults to "amount". :type amount_col: str **Attributes** .. list-table:: :widths: 15 85 :header-rows: 0 * - **expected_range** - None * - **greater_is_better** - False * - **is_bounded_like_percentage** - True * - **name** - Fraud Cost * - **perfect_score** - 0.0 * - **problem_types** - [ProblemTypes.BINARY, ProblemTypes.TIME_SERIES_BINARY] * - **score_needs_proba** - False **Methods** .. autoapisummary:: :nosignatures: evalml.objectives.fraud_cost.FraudCost.calculate_percent_difference evalml.objectives.fraud_cost.FraudCost.can_optimize_threshold evalml.objectives.fraud_cost.FraudCost.decision_function evalml.objectives.fraud_cost.FraudCost.is_defined_for_problem_type evalml.objectives.fraud_cost.FraudCost.objective_function evalml.objectives.fraud_cost.FraudCost.optimize_threshold evalml.objectives.fraud_cost.FraudCost.positive_only evalml.objectives.fraud_cost.FraudCost.score evalml.objectives.fraud_cost.FraudCost.validate_inputs .. py:method:: calculate_percent_difference(cls, score, baseline_score) :classmethod: Calculate the percent difference between scores. :param score: A score. Output of the score method of this objective. :type score: float :param baseline_score: A score. Output of the score method of this objective. In practice, this is the score achieved on this objective with a baseline estimator. :type baseline_score: float :returns: The percent difference between the scores. Note that for objectives that can be interpreted as percentages, this will be the difference between the reference score and score. For all other objectives, the difference will be normalized by the reference score. :rtype: float .. py:method:: can_optimize_threshold(cls) :property: Returns a boolean determining if we can optimize the binary classification objective threshold. This will be false for any objective that works directly with predicted probabilities, like log loss and AUC. Otherwise, it will be true. :returns: Whether or not an objective can be optimized. :rtype: bool .. py:method:: decision_function(self, ypred_proba, threshold=0.5, X=None) Apply a learned threshold to predicted probabilities to get predicted classes. :param ypred_proba: The classifier's predicted probabilities :type ypred_proba: pd.Series, np.ndarray :param threshold: Threshold used to make a prediction. Defaults to 0.5. :type threshold: float, optional :param X: Any extra columns that are needed from training data. :type X: pd.DataFrame, optional :returns: predictions .. py:method:: is_defined_for_problem_type(cls, problem_type) :classmethod: Returns whether or not an objective is defined for a problem type. .. py:method:: objective_function(self, y_true, y_predicted, X, y_train=None, sample_weight=None) Calculate amount lost to fraud per transaction given predictions, true values, and dataframe with transaction amount. :param y_predicted: Predicted fraud labels. :type y_predicted: pd.Series :param y_true: True fraud labels. :type y_true: pd.Series :param y_train: Ignored. :type y_train: pd.Series :param X: Data with transaction amounts. :type X: pd.DataFrame :param sample_weight: Ignored. :type sample_weight: pd.DataFrame :returns: Amount lost to fraud per transaction. :rtype: float :raises ValueError: If amount_col is not a valid column in the input data. .. py:method:: optimize_threshold(self, ypred_proba, y_true, X=None) Learn a binary classification threshold which optimizes the current objective. :param ypred_proba: The classifier's predicted probabilities :type ypred_proba: pd.Series :param y_true: The ground truth for the predictions. :type y_true: pd.Series :param X: Any extra columns that are needed from training data. :type X: pd.DataFrame, optional :returns: Optimal threshold for this objective. :raises RuntimeError: If objective cannot be optimized. .. py:method:: positive_only(cls) If True, this objective is only valid for positive data. Defaults to False. .. py:method:: score(self, y_true, y_predicted, y_train=None, X=None, sample_weight=None) Returns a numerical score indicating performance based on the differences between the predicted and actual values. :param y_predicted: Predicted values of length [n_samples] :type y_predicted: pd.Series :param y_true: Actual class labels of length [n_samples] :type y_true: pd.Series :param y_train: Observed training values of length [n_samples] :type y_train: pd.Series :param X: Extra data of shape [n_samples, n_features] necessary to calculate score :type X: pd.DataFrame or np.ndarray :param sample_weight: Sample weights used in computing objective value result :type sample_weight: pd.DataFrame or np.ndarray :returns: score .. py:method:: validate_inputs(self, y_true, y_predicted) Validate inputs for scoring.