RidgeCV

• Implementa la regresión Ridge con validación cruzada.

• Por defecto usa leave-on-out cross-validation.

from sklearn.datasets import load_diabetes

X, y = load_diabetes(return_X_y=True)

from sklearn.linear_model import RidgeCV

ridgeCV = RidgeCV(
    # ---------------------------------------------------------------------
    # Regularization strength; must be a positive float. Regularization
    # improves the conditioning of the problem and reduces the variance of
    # the estimates. Larger values specify stronger regularization. Alpha
    # corresponds to 1 / (2C) in other linear models such as
    # LogisticRegression or LinearSVC.
    # alphas=(0.1, 1.0, 10.0),
    alphas=[1e-3, 1e-2, 1e-1, 1],
    # ---------------------------------------------------------------------
    # Whether to fit the intercept for this model.
    fit_intercept=True,
    # ---------------------------------------------------------------------
    # A string (see model evaluation documentation) or a scorer callable
    # object / function with signature scorer(estimator, X, y).
    scoring=None,
    # ---------------------------------------------------------------------
    # Determines the cross-validation splitting strategy. Possible inputs
    # for cv are:
    # * None, to use the efficient Leave-One-Out cross-validation
    # * integer, to specify the number of folds.
    # * CV splitter
    # * An iterable yielding (train, test) splits as arrays of indices.
    cv=None,
    # ---------------------------------------------------------------------
    # Flag indicating which strategy to use when performing Leave-One-Out
    # Cross-Validation.
    # * 'auto' : use 'svd' if n_samples > n_features, otherwise use 'eigen'
    # * 'svd'
    # * 'eigen' : force computation via eigendecomposition of X.X^T
    gcv_mode='auto',
    # ---------------------------------------------------------------------
    # Flag indicating if the cross-validation values corresponding to each
    # alpha should be stored in the cv_values_ attribute (see below). This
    # flag is only compatible with cv=None (i.e. using Leave-One-Out
    # Cross-Validation).
    store_cv_values=False,
    # ---------------------------------------------------------------------
    # Flag indicating whether to optimize the alpha value (picked from the
    # alphas parameter list) for each target separately (for multi-output
    # settings: multiple prediction targets). When set to True, after
    # fitting, the alpha_ attribute will contain a value for each target.
    # When set to False, a single alpha is used for all targets.
    alpha_per_target=False,
)

ridgeClassifierCV.fit(X, y)

ridgeClassifierCV.score(X, y)

0.9630931458699473

ridgeClassifierCV.alpha_

0.01

ridgeClassifierCV.coef_

array([[ 2.70962810e-01, -1.99904473e-02, -2.68869779e-02,
        -5.52130776e-04,  7.51784333e-01,  6.70081562e+00,
        -2.24514478e+00, -3.78660924e+00, -2.34922271e-01,
         3.05972457e-01, -9.50391185e-01, -3.25337487e-02,
         7.39111510e-03,  3.34830382e-03, -5.15606649e+00,
        -1.39236707e+00,  4.94530386e+00, -4.00371830e+00,
        -2.54634379e+00, -1.35630031e-02, -3.49442568e-01,
        -5.24646097e-03,  7.11490969e-03,  1.75379598e-03,
        -4.16225726e+00, -5.32987896e-02, -6.21848378e-01,
        -2.48894530e+00, -9.96733111e-01, -4.46615638e+00]])

ridgeClassifierCV.intercept_

array([4.76062228])