Finite sample differentially private confidence intervals
From MaRDI portal
Abstract: We study the problem of estimating finite sample confidence intervals of the mean of a normal population under the constraint of differential privacy. We consider both the known and unknown variance cases and construct differentially private algorithms to estimate confidence intervals. Crucially, our algorithms guarantee a finite sample coverage, as opposed to an asymptotic coverage. Unlike most previous differentially private algorithms, we do not require the domain of the samples to be bounded. We also prove lower bounds on the expected size of any differentially private confidence set showing that our the parameters are optimal up to polylogarithmic factors.
Recommendations
- Privacy-preserving statistical estimation with optimal convergence rates
- Minimax Optimal Procedures for Locally Private Estimation
- Asymptotically Optimal and Private Statistical Estimation
- The cost of privacy: optimal rates of convergence for parameter estimation with differential privacy
- Differential privacy and robust statistics
Cites work
- A statistical framework for differential privacy
- Differential privacy and robust statistics
- Differential privacy for functions and functional data
- Inference using noisy degrees: differentially private -model and synthetic graphs
- Privacy-preserving statistical estimation with optimal convergence rates
- Testing Statistical Hypotheses
- Theory of Cryptography
- What can we learn privately?
Cited in
(13)- The cost of privacy: optimal rates of convergence for parameter estimation with differential privacy
- Tight lower bound of sparse covariance matrix estimation in the local differential privacy model
- Differentially Private Significance Tests for Regression Coefficients
- A Survey of Differentially Private Regression for Clinical and Epidemiological Research
- Differentially private inference via noisy optimization
- Gaussian differentially private robust mean estimation and inference
- Differentially private medians and interior points for non-pathological data
- Simulation-Based, Finite-Sample Inference for Privatized Data
- A polynomial time, pure differentially private estimator for binary product distributions
- Not all learnable distribution classes are privately learnable
- Mixtures of Gaussians are privately learnable with a polynomial number of samples
- On the sample complexity of privately learning unbounded high-dimensional Gaussians
- Differentially private Assouad, Fano, and Le Cam
This page was built for publication: Finite sample differentially private confidence intervals
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4993312)
