DiscreteRandomSmoother

class scikit_weak.utils.DiscreteRandomSmoother(p_err=0.1, p_incl=1.0, prob_ranges=None, type='set', samples=100, epsilon=0.0)

A class to transform a supervised learning problem into a weakly supervised one, based on random sampling. It currently supports transformation to superset and fuzzy label learning. Note that DiscreteRandomSmoother does not implement the transform method: therefore, usage should be based on calling fit_transform.

Parameters

  • p_err (float, default=0.1) – The probability to include any single wrong label in a sample draw. Should be between 0 and 1.

  • p_incl (float, default=1.0) – The probability to include the correct label in a sample draw. Should be between 0 and 1.

  • prob_ranges (enumerable of float, default=None) – Array of possible membership degrees to be sampled. If not None, overrides both p_err and p_incl

  • type ({'set', 'fuzzy'}, default=set) – Type of weakly supervised labels to transform into

  • samples (int, default=100) – Number of samples to be generated

  • epsilon (float, default=1.0) – Parameter to select the minimum allowed label degree. Only used when type == ‘set’. Should be between 0 and 1

Variables

n_classes (int) – The number of unique classes in y

fit_transform(X, y)

Fit to data, then transform it.

Fits transformer to X and y with optional parameters fit_params and returns a transformed version of X.

Xarray-like of shape (n_samples, n_features)

Input samples.

yarray-like of shape (n_samples,) or (n_samples, n_outputs), default=None

Target values (None for unsupervised transformations).

**fit_paramsdict

Additional fit parameters.

X_newndarray array of shape (n_samples, n_features_new)

Transformed array.