BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Inference and Learning in the Anglican Probabilistic Programming S
 ystem - Jan-Willem van de Meent (Oxford)
DTSTART:20160405T100000Z
DTEND:20160405T110000Z
UID:TALK65534@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:Probabilistic programming systems aim to accelerate iterative 
 development of machine learning approaches by introducing an abstraction b
 oundary: models are defined using a domain-specific language\, and a back 
 end implements generic inference methods for such programs. The aim of thi
 s research endeavor is to do for the domains of data science and artificia
 l intelligence what compiler technologies have done for software developme
 nt: enable practitioners to reason about their models at a higher level of
  abstraction.\n\nIn this talk I will discuss inference strategies employed
  in Anglican\, a probabilistic programing system closely integrated with t
 he language Clojure. Anglican has pioneered inference techniques based on 
 sequential Monte Carlo that apply to programs written in general-purpose l
 anguages that support recursion\, higher-order functions\, and black box d
 eterministic primitives. In addition to strategies for posterior inference
 \, I will discuss extensions to policy search and marginal MAP estimation 
 problems.\n\nBIO \n\nJan-Willem is post-doc in Machine Learning at the Dep
 artment of Engineering Science at Oxford. He works primarily on the Anglic
 an probabilistic programming system\, which he co-created with Frank Wood 
 and David Tolpin. His broader research agenda is to understand how program
 s may be used to define structured and composable models for machine learn
 ing and artificial intelligence. To facilitate this agenda\, he also works
  on inference techniques for probabilistic programs.
LOCATION:Engineering Department\, CBL Room BE-438
END:VEVENT
END:VCALENDAR