empulse.models

The models module contains a collection of cost-sensitive and profit-driven models.

Cost-Sensitive and Value-driven Models

B2BoostClassifier	Gradient boosting model to optimize instance-dependent cost loss for customer churn.
CSBaggingClassifier	Bagging classifier to optimize instance-dependent cost loss.
CSBoostClassifier	Gradient boosting model to optimize instance-dependent cost loss.
CSForestClassifier	Random Forest classifier to optimize instance-dependent cost loss.
CSLogitClassifier	Logistic classifier to optimize instance-dependent cost loss.
CSThresholdClassifier	Classifier that sets the decision threshold to optimize the instance-dependent cost loss.
CSTreeClassifier	Decision tree classifier to optimize instance-dependent cost loss.
RobustCSClassifier	Classifier that fits a cost-sensitive classifier with costs adjusted for outliers.
ProfLogitClassifier	Logistic classifier to optimize profit-driven score.

Bias Mitigation Models

BiasRelabelingClassifier	Classifier which relabels instances during training to remove bias against a subgroup.
BiasResamplingClassifier	Classifier which resamples instances during training to remove bias against a subgroup.
BiasReweighingClassifier	Classifier which reweighs instances during training to remove bias against a subgroup.