skada.metrics.PredictionEntropyScorer
- class skada.metrics.PredictionEntropyScorer(greater_is_better=False, reduction='mean', kwargs=None)[source]
Score based on the entropy of predictions on unsupervised dataset.
See [18] for details.
- Parameters:
- greater_is_betterbool, default=False
Whether scorer is a score function (default), meaning high is good, or a loss function, meaning low is good. In the latter case, the scorer object will sign-flip the outcome of the scorer.
- reduction: str, default='mean'
Specifies the reduction to apply to the entropy values. Must be one of ['none', 'mean', 'sum']. If 'none', the entropy values for each sample are returned ([1]_ method). If 'mean', the mean of the entropy values is returned. If 'sum', the sum of the entropy values is returned.
- Returns:
- entropyfloat or ndarray of floats
If reduction is 'none', then ndarray of shape (n_samples,). Otherwise float.
References
[18]Pietro Morerio et al. Minimal-Entropy correlation alignment for unsupervised deep domain adaptation. ICLR, 2018.
