Differentially Private Quantiles with Smaller Error
Improves the error of private quantile estimation in the central model, with matching lower bounds.
Research
Differential privacy, randomized data structures and high-dimensional search. Every entry links straight to the paper, and to the code where there is code.
Selected 2
Conference papers 6
Piquantε: Private Quantile Estimation in the Two-Server Model
Differentially Private Quantiles with Smaller Error
Lightweight Protocols for Distributed Private Quantile Estimation
Differentially Private Release of Hierarchical Origin/Destination Data with a TopDown Approach
Differentially Private High-Dimensional Approximate Range Counting, Revisited
Locality Sensitive Hashing of Trajectories Under Local Differential Privacy
Journal papers 1
Towards a Fair and Comprehensive Evaluation of Walkable Accessibility and Attractivity in the 15 Min City Scenario Based on Demographic Data
Workshop papers 2
InfTDA: a Simple TopDown Mechanism for Hierarchical Differentially Private Counting Queries
Private Quantile Estimation in the Two Server Model
Projects 1
The Privacy Analysis of the Differentially Private Stochastic Gradient Descent
A survey of the state of the art in differentially private deep learning, written for a PhD course on Information Theoretic Models in Security.
Theses 3
Differentially Private Methods for Data Release, Similarity Search and Quantile Estimation
Problems of Ranking and Dynamics of Complex Bipartite Networks
Solitoni in Fluidi Classici e Quantistici