Learning Privately with Labeled and Unlabeled Examples

Abstract

A private learner is an algorithm that given a sample of labeled individual examples outputs a generalizing hypothesis while preserving the privacy of each individual. In 2008, Kasiviswanathan et al. (FOCS 2008) gave a generic construction of private learners, in which the sample complexity is (generally) higher than what is needed for non-private learners… (More)
DOI: 10.1137/1.9781611973730.32

Topics

Cite this paper

@inproceedings{Beimel2015LearningPW, title={Learning Privately with Labeled and Unlabeled Examples}, author={Amos Beimel and Kobbi Nissim and Uri Stemmer}, booktitle={SODA}, year={2015} }