Learning privately with labeled and unlabeled examples
SODA, pp. 461-477, 2015.
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 no...More
PPT (Upload PPT)