Research additional metrics to add them in fair mango

[MarcBresson]

• Predictive Rate Parity
• Group Benefit Disparity
• Recall Parity
• False Positive Rate Parity

[NacerKROUDIR]

there are two metrics that are already implemented:

• Recall Parity: same as Equal Opportunity Difference.
• False Positive Rate Parity: already used in Equalised Odds without exposing it to the end user.

The other two metrics we are planning to implement are:

• Predictive Rate Parity: Compares Positive Predictive Value (Precision Score)
• Group Benefit Disparity: Compares the rate at which a particular event is predicted to occur within a subgroup compared to the rate at which it actually occurs.