View Document Preview and Link Document Date: 2014-07-01 19:35:05 Open Document File Size: 628,67 KB Share Result on Facebook
City Exponential Sampling / Laplacenoise / / Company LG / Microsoft / / / Facility Boston University’s Hariri Institute / Pennsylvania State University / Stanford University / Harvard University / / IndustryTerm by-product / classic gradient descent algorithm / basic and powerful computational tools / polynomial time algorithm / give algorithms / private algorithms / online optimization / stochastic gradient descent algorithm / generic tool / computing / descent-based algorithms / algorithms / private algorithm / / Organization The Pennsylvania State University / Boston University’s Hariri Institute for Computing / Harvard University / Sloan Foundation / National Science Foundation / Stanford University / Computer Science and Engineering Department / / Person Gradient Descent / Adam Smith / / ProvinceOrState Pennsylvania / / SportsLeague Stanford University / / Technology cryptography / specific algorithm / randomized algorithm / Exponential Sampling-based Algorithms / differentially private algorithms / polynomial time algorithm / 1.1 Contributions We give algorithms / classic gradient descent algorithm / machine learning / descent-based algorithms / stochastic gradient descent algorithm / Huberizationbased algorithm / differentially private algorithm / / SocialTag 
Private Empirical Risk Minimization, Revisited Raef Bassily∗ Adam Smith∗ † Abhradeep Thakurta‡