BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Bayesian network structure learning from uncertain interventions -
  Kevin Murphy\, University of British Columbia
DTSTART:20070814T140000Z
DTEND:20070814T150000Z
UID:TALK7808@talks.cam.ac.uk
CONTACT:Oliver Williams
DESCRIPTION:We show how to learn causal structure from data generated by s
 ystems subject to perturbations (interventions) with unknown effects. Our 
 approach is to treat the interventions\, as well as the other system varia
 bles\, as random variables\, and to learn a joint graph topology using Bay
 esian inference. \nThis is in contrast to the standard approach\, which as
 sumes that the effects of interventions are known. We show that\, on a dat
 aset consisting of protein phosphorylation levels measured under various p
 erturbations\, learning the targets of intervention results in models that
  fit the data better than falsely assuming the interventions have known ta
 rgets. Furthermore\, learning the targets of intervention is useful for su
 ch tasks as drug and disease target discovery\, where we wish to distingui
 sh direct effects from indirect effects. We illustrate the latter by corre
 ctly identifying known targets of genetic mutation in various forms of leu
 kemia using microarray expression data. \nJoint work with Daniel Eaton. \n
LOCATION:Small public lecture room\, Microsoft Research Ltd\, 7 J J Thomso
 n Avenue (Off Madingley Road)\, Cambridge
END:VEVENT
END:VCALENDAR