Difference between revisions of "Differential privacy"

From Simson Garfinkel
Jump to navigationJump to search
m
Line 26: Line 26:
=== Improving query accuracy within the privacy budget ===
=== Improving query accuracy within the privacy budget ===


* [https://people.cs.umass.edu/~miklau/assets/pubs/dp/Li15matrix.pdf The matrix mechanism: optimizing linear counting queries
* [https://people.cs.umass.edu/~miklau/assets/pubs/dp/Li15matrix.pdf The matrix mechanism: optimizing linear counting queries under differential privacy], Gerome Miklau, Michael Hay, Andrew McGregor, Vibhor Rastogi,The VLDB Journal, August 2015, DOI 10.1007/s00778-015-0398-x.
under differential privacy], Gerome Miklau, Michael Hay, Andrew McGregor, Vibhor Rastogi,The VLDB Journal, August 2015, DOI 10.1007/s00778-015-0398-x.


=== Differential Privacy and Floating Point Accuracy ===
=== Differential Privacy and Floating Point Accuracy ===

Revision as of 10:34, 14 July 2017

A few references on Differential Privacy, for people who don't want to get bogged down with the math.

Video

Applications

Advanced Topics

Improving query accuracy within the privacy budget

Differential Privacy and Floating Point Accuracy

Floating point math on computer's isn't continuous, and differential privacy implementations that assume it is may experience a variety of errors that result in privacy loss. A discussion of the problems inherently in floating-point arithmetic can be found in Oracle's What Every Computer Scientist Should Know About Floating-Point Arithmetic, an edited reprint of the paper What Every Computer Scientist Should Know About Floating-Point Arithmetic, by David Goldberg, published in the March, 1991 issue of Computing Surveys.

"How Will Statistical Agencies Operate When All Data Are Private?" (MS #1142) has been published to Journal of Privacy and Confidentiality. http://repository.cmu.edu/jpc/vol7/iss3/1

Differential Privacy and the Statistical Agencies


The Fool's Gold Controversy

What's wrong with this article and with the followups?

Other attacks

Math

p for randomized response rate:

$p = \frac{e^\epsilon}{1+e^\epsilon}$

Probability that randomized response should be flipped.

See Also