dataeval.metrics.bias¶

Bias metrics check for skewed or imbalanced datasets and incomplete feature representation which may impact model performance.

Output Classes¶

`BalanceOutput`	Output class for `balance()` bias metric.
`CompletenessOutput`	Output from the completeness function.
`CoverageOutput`	Output class for `coverage()` bias metric.
`DiversityOutput`	Output class for `diversity()` bias metric.
`LabelParityOutput`	Output class for `label_parity()` bias metrics.
`ParityOutput`	Output class for `parity()` bias metrics.

`balance`(metadata[, num_neighbors])	Mutual information (MI) between factors (class label, metadata, label/image properties).
`completeness`(embeddings, quantiles)	Calculate the fraction of boxes in a grid defined by quantiles that
`coverage`(embeddings[, radius_type, num_observations, ...])	Class for evaluating coverage and identifying images/samples that are in undercovered regions.
`diversity`(metadata[, method])	Compute diversity and classwise diversity for discrete/categorical variables through standard histogram binning, for continuous variables.
`label_parity`(expected_labels, observed_labels[, ...])	Calculate the chi-square statistic to assess the parity between expected and observed label distributions.
`parity`(metadata)	Calculate chi-square statistics to assess the linear relationship between multiple factors and class labels.