BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Probabilistic Programming - James Lloyd (University of Cambridge)
DTSTART:20130314T150000Z
DTEND:20130314T163000Z
UID:TALK43632@talks.cam.ac.uk
CONTACT:Colorado Reed
DESCRIPTION:What is the most general class of statistical models? And how 
 can we perform inference without designing custom algorithms? Just as auto
 mated inference algorithms make working with graphical models easy (e.g. B
 UGS)\, a new class of automated inference procedures is being developed fo
 r the more general case of Turing-complete generative models. In this tuto
 rial\, we introduce the practice of specifying generative models as progra
 ms which produce a stochastic output\, and then automatically performing i
 nference on the execution trace of that program\, conditioned on the algor
 ithm having produced a specific output. We give examples of how to specify
  complex models and run inference in several ways\, including recent advan
 ces in automatic Hamiltonian Monte Carlo and variational inference.\n\nRea
 ders who wish to familiarize themselves with these ideas beforehand are en
 couraged to browse around at:\nhttp://probabilistic-programming.org/wiki/H
 ome\nand\nhttp://projects.csail.mit.edu/church/wiki/Church
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR