[Pl-seminar] 11/21: Ryan Culpepper, Introduction to Probabilistic Programming

[Vincent St-Amour]

NUPRL Seminar presents

Ryan Culpepper
NUPRL

3:00 - 4:30 (non-standard time)
Friday, 11/21
Room 366 WVH (http://www.ccs.neu.edu/home/wand/directions.html)


Introduction to Probabilistic Programming

Abstract:
Probabilistic programming languages allow programmers to build
generative probabilistic models using standard programming idioms
plus two features: sampling and conditioning. A wide variety of
machine learning problems can be formulated as inference problems
for probabilistic models conditioned on observed data.

In this talk I will give an introduction to many of the relevant
concepts in probabilistic machine learning. These will include common
structures of probabilistic models, the connection between generative
models and probabilistic programs, and sampling-based inference
techniques. I will use a sketch of a semantics for a core
probabilistic language to discuss some open problems.