Students will discuss their summer research project working with Adam Groce and Andrew Bray:

Often we want to perform data analysis on data sets that contain private, sensitive information. Differentially private data analysis gives a way to perform analysis of this sort of data while proving that the output does not violate anyone's privacy. But coming up with a differentially private way to do any particular analysis can be hard. This talk will discuss new ways to perform hypothesis tests and change point detection while maintaining privacy.