Event Date
SPEAKER: Matthew Reimherr; Assistant Professor, Dept of Statistics, Pennsylvania State University (visiting UC Berkeley)
TITLE: “Differential Privacy for Functional Summaries via the Exponential Mechanism with Applications to Functional PCA”
ABSTRACT: The exponential mechanism is a fundamental tool of Differential Privacy (DP) due to its strong privacy guarantees and flexibility. In this talk I will present a recent extension (joint work with Jordan Awan, Ana Kenney, and Aleksandra Slavkovic from PSU) to settings with summaries based on infinite-dimensional outputs such as with functional data analysis, shape analysis, and nonparametric statistics. We show that one can design the mechanism with respect to a specific base measure over the output space, such as a Gaussian process. We provide a positive result that establishes a Central Limit Theorem for the exponential mechanism quite broadly. We also provide an apparent negative result, showing that the magnitude of the noise introduced for privacy is asymptotically non-negligible relative to the statistical estimation error. We develop an epsilon-DP mechanism for functional principal component analysis, applicable in separable Hilbert spaces. We demonstrate its performance via simulations and applications to two datasets.
*This talk will be tailored to a statistics audience assuming little to no background in data privacy.
SPEAKER'S WEB PAGE: http://www.personal.psu.edu/mlr36/
DATE: Thursday, February 28th, 4:10pm
LOCATION: MSB 2112 (Math Dept, 2nd floor)
REFRESHMENTS: 3:30pm MSB 4110 (4th floor lounge)
STA 290 Seminar List: https://statistics.ucdavis.edu/seminars