skada.KLIEPReweight

skada.KLIEPReweight(base_estimator=None, gamma=1.0, cv=5, n_centers=100, tol=1e-06, max_iter=1000, random_state=None)[source]

KLIEPReweight pipeline adapter and estimator.

see [3] for details.

Parameters:

base_estimatorsklearn estimator, default=LogisticRegression(): estimator used for fitting and prediction
gammafloat or array like: Parameters for the kernels. If array like, compute the likelihood cross validation to choose the best parameters for the RBF kernel. If float, solve the optimization for the given kernel parameter.
cvint, cross-validation generator or an iterable, default=5: Determines the cross-validation splitting strategy. If it is an int it is the number of folds for the cross validation.
n_centersint, default=100: Number of kernel centers defining their number.
tolfloat, default=1e-6: Tolerance for the stopping criterion in the optimization.
max_iterint, default=1000: Number of maximum iteration before stopping the optimization.
random_stateint, RandomState instance or None, default=None: Determines random number generation for dataset creation. Pass an int for reproducible output across multiple function calls.

Returns:

pipelinesklearn pipeline: Pipeline containing the KLIEPReweight adapter and the base estimator.

References

[3]

Masashi Sugiyama et. al. Direct Importance Estimation with Model Selection and Its Application to Covariate Shift Adaptation. In NeurIPS, 2007.

Examples using `skada.KLIEPReweight`