BEGIN:VCALENDAR VERSION:2.0 PRODID:-//wp-events-plugin.com//7.2.3.1//EN TZID:America/New_York X-WR-TIMEZONE:America/New_York BEGIN:VEVENT UID:636@fds.yale.edu DTSTART;TZID=America/New_York:20240930T160000 DTEND;TZID=America/New_York:20240930T170000 DTSTAMP:20250916T142139Z URL:https://fds.yale.edu/events/sds-seminar-adam-smith-bu/ SUMMARY:S&\;DS Seminar: Adam Smith (BU)\, "Privacy in Machine Learning a nd Statistical Inference" DESCRIPTION:Zoom Link: https://yale.zoom.us/j/94223816617 Meeting ID: 942 2 381 6617\n\n\n\nAbstract: The results of learning and statistical inferen ce reveal information about the data they use. This talk discusses the pos sibilities and limitations of fitting machine learning and statistical mod els while protecting the privacy of individual records.\n\n\n\nI will begi n by explaining what makes this problem difficult\, using recent work on e xample memorization as an illustration. I will then present differential p rivacy\, a rigorous definition of privacy in statistical databases that is now widely studied\, and increasingly used to analyze and design deployed systems. \n\n\n\nFinally\, I will present recent algorithmic results on two fundamental problems: differentially private mean estimation and linea r regression. We give time- and sample-efficient algorithms for “nicely ” distributed (e.g. subgaussian) data. These algorithms adapt automatica lly to the geometry of the instance at hand\, yielding instance-optimal ac curacy guarantees.\n\n\n\nThe talk will be based on joint work with (amon g others) Gavin Brown\, Mark Bun\, Vitaly Feldman\, Sam Hopkins\, and Kuna l Talwar (arxiv 2012.06421\, 2301.12250\, 2404.15409).\n\n\n\nBio: Adam Sm ith is a Professor of Computer Science and Engineering at Boston Universit y. His research interests lie in data privacy and cryptography\, and their connections to machine learning\, statistics\, information theory\, and q uantum computing. He obtained his Ph.D. from MIT in 2004 and has held visi ting positions at the Weizmann Institute of Science\, UCLA\, and Harvard. He previously was a Professor of Computer Science and Engineering at Penn State. He received a Presidential Early Career Award for Scientists and En gineers (PECASE) in 2009\; a 2016 Theory of Cryptography Test of Time awar d\; the 2019 Eurocrypt Test of Time award\; and the 2017 Gödel Prize.\n\n \n\nLink to personal site: http://cs-people.bu.edu/ads22/\n CATEGORIES:FDS Events,Statistics & Data Science Seminar END:VEVENT BEGIN:VTIMEZONE TZID:America/New_York X-LIC-LOCATION:America/New_York BEGIN:DAYLIGHT DTSTART:20240310T030000 TZOFFSETFROM:-0500 TZOFFSETTO:-0400 TZNAME:EDT END:DAYLIGHT END:VTIMEZONE END:VCALENDAR