Moving beyond data cleaning, students work with algorithms that incorporate fairness constraints into the loss function (adversarial debiasing) and adjust decision thresholds after training. They compare the computational cost and efficacy of these methods.

Similar Lessons