The model of differential privacy has become widely accepted as a suitable way to release information about private data while preserving the privacy of the individuals contributing to the data. One reason for its popularity is the availability of several “mechanisms” that produce output meeting the differential privacy guarantee. A second reason is that there are simple rules for reasoning about the composition of multiple differentially private outputs. Taken together, this provides a set of building blocks and cement which can be used to produce algorithms for private data release. In this tutorial, I will review some of the popular mechanisms for differential privacy, and show some of the ways that they can be used for a variety of different data releases
This file was generated by bibtex2html 1.92.