BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Completely Random Measures in Bayesian Nonparametrics - Dr Daniel 
 Roy (University of Cambridge)\, Creighton Heaukulani
DTSTART:20121018T133000Z
DTEND:20121018T150000Z
UID:TALK41135@talks.cam.ac.uk
CONTACT:Konstantina Palla
DESCRIPTION:In Bayesian nonparametric modelling\, we allow a model to lear
 n an unbounded number of parameters by replacing classical finite-dimensio
 nal \nstatistical distributions with infinite-dimensional stochastic proce
 sses.  Completely random measures are a special class of stochastic proces
 ses\, \nwhich are intuitive\, highly interpretable\, and useful for statis
 tical applications.  While this theory is mathematically deep\, in this tu
 torial we will instead \nfocus on practicalities such as how to sample suc
 h objects\, providing pointers to more theoretical material when necessary
 .  This tutorial will include an \nintroduction to random measures and the
 ir practical uses\, Poisson processes and how to sample them\, complete ra
 ndomness\, and how to characterise \na completely random measure.  Finally
 \, we introduce some important completely random measures\, such as the Be
 rnoulli process\, the Beta process\, \nand the Gamma process\, and we will
  show how they are related to some popular objects used in machine learnin
 g\, such as the Dirichlet process\, \nChinese restaurant process and the I
 ndian buffet process.
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR