What are the three fundamental fairness criteria discussed in the chapter for ensuring algorithmic fairness in financial services?
The three fundamental fairness criteria discussed are demographic parity, equal opportunity, and disparate impact.
The chapter discusses three fundamental fairness criteria for ensuring algorithmic fairness in financial services: demographic parity, equal opportunity, and disparate impact. Demographic parity requires that decision outcomes be independent of protected attributes, aiming for equal acceptance rates across groups. Equal opportunity focuses on ensuring equal true positive rates across groups, meaning equally qualified candidates should have equal chances of positive outcomes regardless of protected attributes. Disparate impact originates from legal standards and prohibits practices that disproportionately harm protected groups, using the "80% rule" as a benchmark for assessing discrimination.
Key points
- Demographic parity ensures decision outcomes are independent of protected attributes.
- Equal opportunity ensures equal true positive rates across groups.
- Disparate impact prohibits practices that disproportionately harm protected groups.
- Demographic parity may ignore legitimate differences in qualifications.
- Equal opportunity requires unbiased ground truth labels.
- Disparate impact uses the "80% rule" to assess discrimination.
Related questions
AI and ML Governance in Financial Services
Richard Gwashy Young
Routledge, Taylor & Francis Group