This page refers to the Spring 2017 offering of CSE 712 only. The information on this page does not necessarily apply to every offering of CSE 712.
Research and issues and hazards in inference from large data sets; statistical analysis; modeling.
Similar to last year's seminar on large data but with a different set of featured topics (last year emphasized framing hypotheses for investigation and the Bootstrap technique for significance testing and verification of theoretical error bars---this too will be summarized and updated): 1. Issues and surprising properties presented by noise in the data. This has emerged as an explanation of phenomena in my recent articles https://rjlipton.wordpress.com/2016/11/30/when-data-serves-turkey/ and https://rjlipton.wordpress.com/2016/12/08/magnus-and-the-turkey-grinder/ 2. Handling fitting landscapes that do not have unique global minima. This will include hands-on experiments with the program described in my article https://rjlipton.wordpress.com/2016/11/08/unskewing-the-election/ 3. Related topics of students' choosing; this can involve topics applying ideas from machine learning courses.
M.S.: This course does not fulfill core area (depth) or core course (breadth) requirements.