Not all learnable distribution classes are privately learnable

MaRDI QIDQ7017108zbMATH OpenFDO

Authors Mark Bun, Gautam Kamath, Vikrant Singhal

Publication date 6 March 2025

Full work available at URL https://proceedings.mlr.press/v237/bun24b.html

zbMATH Keywords

differential privacydistribution learninglower bounds

Mathematics Subject Classification ID

Privacy of data (68P27) Learning and adaptive systems in artificial intelligence (68T05)

Cites work

• An equivalence between private classification and online prediction
• Bounds on the sample complexity for private learning and private data release
• Combinatorial methods in density estimation
• Differentially private Assouad, Fano, and Le Cam
• Differentially private release and learning of threshold functions
• Finite sample differentially private confidence intervals
• scientific article; zbMATH DE number 5485439 (Why is no real title available?)
• Learning Poisson binomial distributions
• Near-optimal Sample Complexity Bounds for Robust Learning of Gaussian Mixtures via Compression Schemes
• On the learnability of discrete distributions
• On the sample complexity of privately learning unbounded high-dimensional Gaussians
• Private PAC learning implies finite Littlestone dimension
• Sample complexity bounds on differentially private learning via communication complexity
• The cost of privacy: optimal rates of convergence for parameter estimation with differential privacy
• Theory of Cryptography

This page was built for publication: Not all learnable distribution classes are privately learnable

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7017108)