An axiomatic Positivstellensatz for abstract function algebras over ordered fields, with applications to neural network optimality certificates.
Kearns and Seth Neel and Aaron Roth and Logan Stapleton and Zhiwei Steven Wu , title =
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
abstract
We consider settings in which the right notion of fairness is not captured by simple mathematical definitions (such as equality of error rates across groups), but might be more complex and nuanced and thus require elicitation from individual or collective stakeholders. We introduce a framework in which pairs of individuals can be identified as requiring (approximately) equal treatment under a learned model, or requiring ordered treatment such as "applicant Alice should be at least as likely to receive a loan as applicant Bob". We provide a provably convergent and oracle efficient algorithm for learning the most accurate model subject to the elicited fairness constraints, and prove generalization bounds for both accuracy and fairness. This algorithm can also combine the elicited constraints with traditional statistical fairness notions, thus "correcting" or modifying the latter by the former. We report preliminary findings of a behavioral study of our framework using human-subject fairness constraints elicited on the COMPAS criminal recidivism dataset.
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
Face-Feature Tuning is a label-free logit remapping method that reduces FPR/TPR gaps across groups in deepfake detection while preserving overall accuracy.
Monotonic fairness, enforced by positive-weight neural networks, avoids both class and score resentment by construction while still allowing a demographic parity trade-off.
A fairness-enhanced sampling framework that pseudo-labels unlabeled data, re-samples group-balanced datasets, and bagging ensembles base models, showing reduced demographic disparity in several experiments.
Spectral initialization plus gradient descent recovers a Mahalanobis metric from Bradley-Terry triplet queries with convergence guarantees, and fairness under the estimate transfers to the true metric.
citing papers explorer
-
When Can One Obtain Certificates of Optimality Using Positivstellensaetze?
An axiomatic Positivstellensatz for abstract function algebras over ordered fields, with applications to neural network optimality certificates.
-
Toward Calibrated, Fair, and accurate Deepfake Detection
Face-Feature Tuning is a label-free logit remapping method that reduces FPR/TPR gaps across groups in deepfake detection while preserving overall accuracy.
-
Avoiding Resentment Via Monotonic Fairness
Monotonic fairness, enforced by positive-weight neural networks, avoids both class and score resentment by construction while still allowing a demographic parity trade-off.
-
Fairness in Semi-supervised Learning: Unlabeled Data Help to Reduce Discrimination
A fairness-enhanced sampling framework that pseudo-labels unlabeled data, re-samples group-balanced datasets, and bagging ensembles base models, showing reduced demographic disparity in several experiments.
-
Operationalizing Individual Fairness via Gradient Descent and Bradley-Terry Models
Spectral initialization plus gradient descent recovers a Mahalanobis metric from Bradley-Terry triplet queries with convergence guarantees, and fairness under the estimate transfers to the true metric.