Fairness for machine learning

Hello everybody,

I detected a general model to be biased by gender and I wanted to demonstrate it with fairness methods the same I did manually by intuition by separating the models by gender. My predictor variables were risk factors such a daily habits, exercise etc but the gender variable was included. And I don´t want to add the gender predictor in my model. But I saw in the manual of matlab that they include the predictors that we want to check for fairness in the model. So my question is should I repeat my models with gender to be able to compute the statistical parity difference by gender? I also want to try by age. So should I include both?

Thank you very much

[Sudarshan]

Could you provide more information on which manual you are referring to? Also, do provide some more details or links about your specific problem.