Differential Privacy with Compression
- Creators
- Zhou, Shuheng
- Ligett, Katrina
- Wasserman, Larry
Abstract
This work studies formal utility and privacy guarantees for a simple multiplicative database transformation, where the data are compressed by a random linear or affine transformation, reducing the number of data records substantially, while preserving the number of original input variables. We provide an analysis framework inspired by a recent concept known as differential privacy (Dwork 06). Our goal is to show that, despite the general difficulty of achieving the differential privacy guarantee, it is possible to publish synthetic data that are useful for a number of common statistical learning applications. This includes high dimensional sparse regression (Zhou et al. 07), principal component analysis (PCA), and other statistical measures (Liu et al. 06) based on the covariance of the initial data.
Additional Information
We thank Avrim Blum and John Lafferty for helpful discussions. KL is supported in part by an NSF Graduate Research Fellowship. LW and SZ's research is supported by NSF grant CCF-0625879, a Google research grant and a grant from Carnegie Mellon's Cylab.Attached Files
Submitted - 0901.1365.pdf
Files
Name | Size | Download all |
---|---|---|
md5:b5f56434b456e0ebddf52d9e55e34b78
|
149.5 kB | Preview Download |
Additional details
- Eprint ID
- 96884
- Resolver ID
- CaltechAUTHORS:20190702-110751042
- NSF Graduate Research Fellowship
- NSF
- CCF-0625879
- Carnegie Mellon University
- Created
-
2019-07-08Created from EPrint's datestamp field
- Updated
-
2023-06-02Created from EPrint's last_modified field