On the Robustness of Information-Theoretic Privacy Measures and Mechanisms
Consider a data publishing setting for a dataset composed by both private and non-private features. The publisher uses an empirical distribution, estimated from $n$ i.i.d. samples, to design a privacy mechanism which is applied to new fresh samples afterward. In this paper, we study the discrepancy...
Αποθηκεύτηκε σε:
| Κύριοι συγγραφείς: | , , , |
|---|---|
| Μορφή: | Άρθρο |
| Έκδοση: |
2018
|
| Θέματα: | |
| Διαθέσιμο Online: | https://arxiv.org/abs/1811.06057 http://arxiv.org/abs/1811.06057 |
| Ετικέτες: |
Δεν υπάρχουν, Καταχωρήστε ετικέτα πρώτοι!
|