BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Structure in tensor-variate data: a trivial byproduct of simpler p
 henomena? - John P. Cunningham
DTSTART:20180306T110000Z
DTEND:20180306T120000Z
UID:TALK102463@talks.cam.ac.uk
CONTACT:Dr R.E. Turner
DESCRIPTION:As large tensor-variate data become increasingly common across
  machine learning and statistics\, complex analysis methods for these data
  similarly increase in prevalence.  Such a trend offers the opportunity to
  understand subtler and more meaningful features of the data that\, ostens
 ibly\, could not be studied with simpler datasets or simpler methodologies
 .  While promising\, these advances are also perilous: novel analysis tech
 niques do not always consider the possibility that their results are in fa
 ct an expected consequence of some simpler\, already-known feature of simp
 ler data.  For example\, suppose one fits a time series model (e.g. Kalman
  Filter or multivariate GARCH) to data indexed by time\, measurement dimen
 sion\, and experimental sample.  Was a particular model fit achieved simpl
 y because the data was temporally smooth\, and/or had correlated dimension
 s (or samples)?  I will present two works that address this growing proble
 m\, the first of which uses Kronecker algebra to derive a tensor-variate m
 aximum entropy distribution that has user-specified moments along each mod
 e.  This distribution forms the basis of a statistical hypothesis test\, a
 nd I will use this test to answer two active debates in the neuroscience c
 ommunity over the triviality of certain observed structure in data.  In th
 e second part\, I will discuss how to extend this maximum entropy formulat
 ion to arbitrary constraints using deep neural network architectures in th
 e flavor of implicit generative modeling\, and I will use this method in a
  texture synthesis application.\n\nJohn P. Cunningham is an associate prof
 essor in the Department of Statistics at Columbia University. He received 
 a B.A. in computer science from Dartmouth College\, and a M.S. and Ph.D. i
 n electrical engineering from Stanford University\, and he completed postd
 octoral work in the Machine Learning Group at the University of Cambridge.
  His research group at Columbia investigates several areas of machine lear
 ning and statistical neuroscience.  http://stat.columbia.edu/~cunningham/
LOCATION:CBL Seminar Room
END:VEVENT
END:VCALENDAR