sparse_ho.criterion.CrossVal¶
- class sparse_ho.criterion.CrossVal(criterion, cv=None)¶
Cross-validation loss.
- Parameters
- criterioninstance of
BaseCriterion A criterion that follows the sparse-ho API.
- cvint, cross-validation generator or iterable, default=None
Determines the cross-validation splitting strategy. Possible inputs for cv are:
None, to use the default 5-fold cross-validation,
int, to specify the number of folds.
scikit-learn CV splitter
An iterable yielding (train, test) splits as arrays of indices.
For int/None inputs, KFold is used.
- criterioninstance of
- Attributes
- dict_critsdict
The instances of criterion used for each fold.
- __init__(criterion, cv=None)¶
Methods
__init__(criterion[, cv])get_val(model, X, y, log_alpha[, monitor, tol])Get value of criterion.
get_val_grad(model, X, y, log_alpha, ...[, ...])Get value and gradient of criterion.
get_val_outer(*args, **kwargs)Get value of outer criterion.
proj_hyperparam(model, X, y, log_alpha)Project hyperparameter on a range of admissible values.