Differential Privacy: A Survey of Tools and Its Application to Classification

Differential privacy is a privacy definition designed by
cryptographers which has gained significant attention in the past few
years. This talk will provide a brief tutorial on differential
privacy. We will begin with the basic privacy definitions and their
variants, and then talk about two different methods for designing
differentially private algorithms -- the sensitivity method and the
exponential mechanism. Finally, we will talk about the application of
differential privacy to classification, and provide a survey of some
known results.

This talk will be more informal than the usual AI seminars, and will
be a chalk-board talk.