diffpriv

An implementation of major general-purpose mechanisms for privatizing statistics, models, and machine learners, within the framework of differential privacy of Dwork et al. (2006) doi:10.1007/11681878_14. Example mechanisms include the Laplace mechanism for releasing numeric aggregates, and the exponential mechanism for releasing set elements. A sensitivity sampler (Rubinstein & Alda, 2017) permits sampling target non-private function sensitivity; combined with the generic mechanisms, it permits turn-key privatization of arbitrary programs.