Privacy, Stability and High-dimensional Sparse Regression
Theory of Computation Seminar
Adam Smith, Penn State and Boston University
Tuesday, November 5, 2013 - 1:30pm to 2:30pm
Pierce Hall 320
I will discuss recent results on how different notions of stability of learning algorithms can be used to design differentially private algorithms. We focus on designing algorithms for statistical modelselection. Given a data set and a discrete collection of models, each of which is a family of probability distributions, the goal is to...
Read more about Nov 5: "Privacy, Stability and High-dimensional Sparse Regression" (Adam Smith, Penn State and Boston University)