BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Factor Analysis with a Mixture of Gaussian Factors\, with Applicat
 ion to Separation of the Cosmic Microwave Background - Prof Simon Wilson\,
  School of Computer Science and Statistics\, University College\, Dublin
DTSTART:20090610T131500Z
DTEND:20090610T141500Z
UID:TALK16942@talks.cam.ac.uk
CONTACT:Rachel Fogg
DESCRIPTION:We describe an approach to factor analysis/source separation f
 or multi-channel image data where the sources are assumed to be Gaussian m
 ixtures\, which may be independent or dependent e.g. mixtures of multivari
 ate Gaussians. An MCMC procedure has been developed that implements a full
 y Bayesian procedure e.g. it computes the posterior distribution of source
 s\, their Gaussian mixture parameters and the matrix of linear coefficient
 s from the data.  The method is applied to recovery of the cosmic microwav
 e background \nCMB)\, being an example of source separation applied to ima
 ge data. The CMB is one of many sources of extraterrestrial microwave radi
 ation and we observe a weighted sum of these sources from the Earth at dif
 ferent frequencies. Its accurate reconstruction is of great interest to as
 tronomers and physicists since knowledge of its properties\, and in partic
 ular its anisotropies\, will place strong restrictions on current cosmolog
 ical theories.  From the perspective of a Bayesian solution\, this applica
 tion is interesting as there is considerable prior information about the l
 inear coefficients and the sources.\nResults from  the analysis of data fr
 om the WMAP satellite will be presented\, where microwave radiation is obs
 erved at 5 frequencies and separated into sources\, including the CMB. A d
 iscussion of the many outstanding issues in this problem is also presented
 .
LOCATION:LR11\, Engineering\, Department of
END:VEVENT
END:VCALENDAR