Algorithmic Bias Correction

Fair feature masking

Processing strategy that selectively masks or transforms potentially biased features during training to force the model to rely on non-discriminatory attributes.

← Zurück