Closed jaared closed 1 year ago
Is correlation remover method a good option to mitigate bias when using the statistical parity ratio metric?
That is a very good question and I had to think about it. The answer is no.
The correlation remover should help meet the separation fairness criteria. It should improve these fairness metrics for binary classification:
Correlation remover doesn't show up in plots when using compare_mitigation_methods()
Suggest we add it to the list here.
https://github.com/EqualityAI/EqualityML/blob/90e04435007a653b5c4c69dc5a9b86e0c5d34ce7/equalityml/fair.py#L172-L182