ML Model Bias & Fairness Score

Quantify demographic parity difference, equalized odds, and disparate impact ratio to audit ML model fairness across protected groups

Each component has a specific meaning:

• Protected groups — The protected groups recorded for the patient or scenario being assessed.

Note: Interpret the ml bias score result against the clinical thresholds and context described above.

Enter the protected groups for the patient or scenario you are assessing. Quantify demographic parity difference, equalized odds, and disparate impact ratio to audit ML model fairness across protected groups. Use the ml bias score result to inform your clinical assessment.