

Wasserstein gradient flows and applications to sampling in machine learning - lecture 1
De Anna Korba


Wasserstein gradient flows and applications to sampling in machine learning - lecture 2
De Anna Korba
De Sewoong Oh
Apparaît dans la collection : 2016 - T1 - WS5 - Secrecy and privacy theme
Interactive querying of a database degrades the privacy level. In this paper we answer the fundamental question of characterizing the level of differential privacy degradation as a function of the number of adaptive interactions and the differential privacy levels maintained by the individual queries. Our solution is complete: the privacy degradation guarantee is true for every privacy mechanism and, further, we demonstrate a sequence of privacy mechanisms that do degrade in the characterized manner. The key innovation is the introduction of an operational interpretation (involving hypothesis testing) to differential privacy and the use of the corresponding data processing inequalities. Our result improves over the state of the art and has immediate applications to several problems studied in the literature.