Goto

Collaborating Authors

 dworketal

Envy-Free Classification

Fairness Behind a Veil of Ignorance: A Welfare Analysis for Automated Decision Making

Neural Information Processing Systems

Alsoas expected, smaller values (i.e. See Proposition 4 in Appendix B, whereweformallyprovethis specialcaseoflinearpredictors.

Learningdiscretedistributions: uservsitem-level privacy

Neural Information Processing Systems

An important aspect of the abovedefinition is the notion of neighboring or adjacent datasets.

PrivatelyPublishablePer-instancePrivacy

Neural Information Processing Systems

We consider how to privately share the personalized privacy losses incurred by objective perturbation, using per-instance differential privacy (pDP).

PAC-BayesLearningBoundsfor Sample-DependentPriors

Neural Information Processing Systems

Sample-dependent priors are also relevant in emerging scenarios such as adversarial training. A common practice here is to smooth a given classifier by injecting Gaussian noise into the inputs.