Intersectional Divergence is a new regression fairness measure that combines group error differences with relevance-weighted target values, and it can be optimized as a loss to reduce unfairness.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Intersectional Divergence: Measuring Fairness in Regression
Intersectional Divergence is a new regression fairness measure that combines group error differences with relevance-weighted target values, and it can be optimized as a loss to reduce unfairness.