[Ml-stat-talks] learn about stan from the expert: monday at 2:00pm

David Blei blei at CS.Princeton.EDU
Sat Dec 14 19:34:40 EST 2013

machine learning and statistics aficionados,

michael betancourt is speaking on monday at 2:00PM.  he is one of the
forces behind the stan probabilistic programming system, which lets
anyone easily specify a model and then fit it (okay, approximate the
posterior) using HMC.

this talk is for anyone interested in building customized probability
models to solve their problems.  if you have any doubts, download the
manual at mc-stan.org and you will be counting the hours until monday.

details are below.



Stan: Practical Bayesian Inference with Hamiltonian Monte Carlo
Michael Betancourt
University College London

Monday, December 16
CS Room 302

Practical implementations of Bayesian inference are often limited to
approximation methods that only slowly explore the posterior
distribution.  By taking advantage of the curvature of the posterior,
however, Hamiltonian Monte Carlo (HMC) efficiently explores even the
most highly contorted distributions.  In this talk I will review the
foundations of and recent developments within HMC, concluding with a
discussion of Stan, a powerful inference engine that utilizes HMC,
automatic differentiation, and adaptive methods to minimize user

Bio: Caltech undergrad, PhD from MIT in physics, then a short stint in
industry before working with Andrew Gelman at Columbia and then
starting the postdoc at UCL with Mark Girolami.

More information about the Ml-stat-talks mailing list