Second vignette bernstein on: Bernstein approximations and use of DPMechBernstein for private function release.
Minor edits to docs
diffpriv 0.4.1
Expanding test coverage of Bernstein mechanism and function approximation code.
diffpriv 0.4.0
Addition of S3 constructor and predict()generic implementation for fitting (non-iterated) Bernstein polynomial function approximations.
Addition of DPMechBernstein class implementing the Bernstein mechanism of Alda and Rubinstein (AAAI’2017), for privately releasing functions.
Bug fix in the Laplace random sampler affectingDPMechLaplace
Unit test coverage of new functionality; general documentation improvements.
diffpriv 0.3.2
Addition of DPMechGaussian class for the generic Gaussian mechanism to README, Vignette. Resolves #2
Minor test additions.
diffpriv 0.3.1
Refactoring around releaseResponse() method inDPMechNumeric. Resolves #1
Increased test coverage.
diffpriv 0.3.0
New DPMechGaussian class implementing the Gaussian mechanism, which achieves (epsilon,delta)-differential privacy by adding Gaussian noise to numeric responses calibrated by L2-norm sensitivity.
Refactoring of DPMechGaussian andDPMechLaplace underneath a new VIRTUAL classDPMechNumeric which contains common methods,dims slot (formerly dim changed becausedim is a special slot for S4).
diffpriv 0.2.0
DPMechLaplace objects can now be initialized without specifying non-private target response dim. In such cases, the sensitivity sampler will perform an additionaltarget probe to determine dim.
diffpriv 0.1.0.901
Sensitivity sampler methods no longer require oracles that return lists. Acceptable oracles may now return lists, matrices, data frames, numeric vectors, or char vectors. As a consequence some example code in docs, README and vignette, is simplified.