Glossary term
Glossary term
Governance and Compliance
A fairness metric that is satisfied if the results of a model's classification are not dependent on a given sensitive attribute.
For example, if both Lilliputians and Brobdingnagians apply to Glubbdubdrib University, demographic parity is achieved if the percentage of Lilliputians admitted is the same as the percentage of Brobdingnagians admitted, irrespective of whether one group is on average more qualified than the other.
Contrast with equalized odds and equality of opportunity, which permit classification results in aggregate to depend on sensitive attributes, but don't permit classification results for certain specified ground truth labels to depend on sensitive attributes. See "Attacking discrimination with smarter machine learning" for a visualization exploring the tradeoffs when optimizing for demographic parity.
See Fairness: demographic parity in Machine Learning Crash Course for more information.
For example, if both Lilliputians and Brobdingnagians apply to Glubbdubdrib University, demographic parity is achieved if the percentage of Lilliputians admitted is the same as the percentage of Brobdingnagians admitted, irrespective of whether one group is on average more qualified than the other.
Created for this library
A hiring-tech vendor measures demographic parity in its recruiter ranking model and reports the disparity to enterprise customers as part of its model card.
A credit-risk model owner monitors demographic parity weekly so any sudden divergence can be investigated and explained.
A health-tech startup tracks demographic parity on its triage tool across age groups and reports parity gaps in monthly governance reviews.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License