Skip to content

Latest commit

 

History

History
34 lines (23 loc) · 1.08 KB

README.md

File metadata and controls

34 lines (23 loc) · 1.08 KB
Raw

Pufferfish Privacy Mechanisms for Correlated Data

This repository contains code for our paper "Pufferfish privacy mechanisms for correlated data".

Overview of the code

The code is written in MATLAB.

run_synthetic.m runs MQMApprox, MQMExact and GK16 on a binary Markov chain to compute the noise scale.

run_realData.m runs MQMApprox, MQMExact and GroupDP on a multi-state Markov chain to compute the noise scale. It needs the transition matrix and the Markov chain as input (for example, you may use the power usage AMPds dataset).

The three functions named k_findBest_2dir*.m are the main functions for MQMApprox and MQMExact; and the two functions named inferentialPrivacy*.m are the main functions for GK16.

Citing this work

@inproceedings{song2017pufferfish,
  title={Pufferfish privacy mechanisms for correlated data},
  author={Song, Shuang and Wang, Yizhen and Chaudhuri, Kamalika},
  booktitle={Proceedings of the 2017 ACM International Conference on Management of Data},
  pages={1291--1306},
  year={2017}
}