BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Group invariance and computational sufficiency - Vincent Vu (Ohio 
 State University)
DTSTART:20180629T084500Z
DTEND:20180629T093000Z
UID:TALK107521@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:Statistical sufficiency formalizes the notion of data reductio
 n. In the decision theoretic interpretation\, once a model is chosen all i
 nferences should be based on a sufficient statistic.  However\, suppose we
  start with a set of methods that share a sufficient statistic rather than
  a specific model.  Is it possible to reduce the data beyond the statistic
  and yet still be able to compute all of the methods?  In this talk\, I&#3
 9\;ll present some progress towards a theory of "computational sufficiency
 " and show that strong reductions _can_ be made for large classes of penal
 ized M-estimators by exploiting hidden symmetries in the underlying optimi
 zation problems.  These reductions can (1) enable efficient computation an
 d (2) reveal hidden connections between seemingly disparate methods.  As a
  main example\, I&#39\;ll show how the theory provides a surprising answer
  to the following question: "What do the Graphical Lasso\, sparse PCA\, si
 ngle-linkage clustering\, and L1 penalized Ising model selection all have 
 in common?"
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR