BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:EE-SRI: Prediction of correlation structure from large random matr
 ices
DTSTART:20120621T161500
DTEND:20120621T171500
DTSTAMP:20260929T090437Z
UID:3c582f05fdfedb4e1d597d691aaab5c8dc7663b1c9fd3cfab0252002
CATEGORIES:Conferences - Seminars
DESCRIPTION:Alfred Hero\, University of Michigan\, Ann Arbor\nAbstract\nRa
 ndom matrices arise in many areas of engineering\, social sciences\, and n
 atural sciences. For example\, when rows of the random matrix record succe
 ssive samples of a multivariate response the sample correlation between th
 e columns can reveal important dependency structure in the multivariate re
 sponse\, e.g.\, stars\, hubs and triangles of co-dependency. However\, whe
 n the number of samples is finite and the number p of columns increases su
 ch exploration becomes futile due to a phase transition phenomenon: spurio
 us discoveries will eventually dominate. In this presentation I will prese
 nt theory for predicting these phase transitions and present Poisson limit
  theorems that can be used to predict finite sample behavior of correlatio
 n structure. We will discuss an application to longitudinal gene expressio
 n analysis.
LOCATION:Room MXF1
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR