A classifier decision is fair if it has a fair explanation (prime implicant without protected features, respecting constraints); the paper relates three such fairness notions for classifiers and studies the complexity of testing them.
Fairness in criminal justice risk assessments: The state of the art.Sociological Methods & Research, 50(1):3–44, 2021
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
verdicts
UNVERDICTED 3representative citing papers
Introduces learning-augmented robust algorithmic recourse that trades off consistency with accurate future-model predictions against robustness to inaccurate predictions via a novel algorithm.
Short-term group fairness in repeated selections can incur a high price of fairness even with nearly identical group distributions, but long-term disparities can vanish under simple investment policies with low PoF.
citing papers explorer
-
Fairness of Classifiers in the Presence of Constraints between Features
A classifier decision is fair if it has a fair explanation (prime implicant without protected features, respecting constraints); the paper relates three such fairness notions for classifiers and studies the complexity of testing them.
-
Learning-Augmented Robust Algorithmic Recourse
Introduces learning-augmented robust algorithmic recourse that trades off consistency with accurate future-model predictions against robustness to inaccurate predictions via a novel algorithm.
-
Price of Fairness in Short-Term and Long-Term Algorithmic Selections
Short-term group fairness in repeated selections can incur a high price of fairness even with nearly identical group distributions, but long-term disparities can vanish under simple investment policies with low PoF.