Halifax, NS, Stan talk and course Thu 19 Oct

[This article was first published on R – Statistical Modeling, Causal Inference, and Social Science, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Halfiax, here we come.

I (Bob, not Andrew) am going to be giving a talk on Stan and then Mitzi and I will be teaching a course on Stan after that. The public is invited, though space is limited for the course. Here are details if you happen to be in the Maritime provinces.

TALK: Stan: A Probabilistic Programming Language for Bayesian Inference

Date: Thursday October 19, 2017

Time: 10am

Location: Slonim Conference room (#430), Goldberg Computer Science Building, Dalhousie University, 6050 University Avenue, Halifax

Abstract

I’ll describe Stan’s probabilistic programming language, and how it’s used, including

  • blocks for data, parameter, and predictive quantities
  • transforms of constrained parameters to unconstrained spaces, with automatic Jacobian corrections
  • automatic computation of first- and higher-order derivatives
  • operator, function, and linear algebra library
  • vectorized density functions, cumulative distributions, and random number generators
  • user-defined functions
  • (stiff) ordinary differential equation solvers

I’ll also provide an overview of the underlying algorithms for full Bayesian inference and for maximum likelihood estimation:

  • adaptive Hamiltonian Monte Carlo for MCMC
  • L-BFGS optimization and transforms for MLE

I’ll also briefly describe the user-facing interfaces: RStan (R), PyStan (Python), CmdStan (command line), Stan.jl (Julia), MatlabStan (MATLAB)

I’ll finish with an overview of the what’s on the immediate horizon:

  • GPU matrix operations
  • MPI multi-core, multi-machine parallelism
  • data parallel expectation propagation for approximate Bayes
  • marginal Laplace approximations

TUTORIAL: Introductio to Bayesian Modeling and Inference with RStan

Instructors:

  • Bob Carpenter, Columbia University
  • Mitzi Morris, Columbia University

Date: Thursday October 19, 2017

Time: 11:30am-5:30pm (following the seminar on Stan at 10am)

Location: Slonim Conference room (#430)
Goldberg Computer Science Building
Dalhousie University
6050 University Avenue, Halifax

Registration: EventBrite Registration Page

Description:

This short course will provide

  • an introduction to Bayesian modeling
  • an introduction to Monte Carlo methods for Bayesian inference
  • an overview of the probabilistic programming language Stan

Stan provides a language for coding Bayesian models along
with state-of-the-art inference algorithms based on gradients. There will be an overview of how Stan works, but the main focus will be on the RStan interface and building applied models.

The afternoon will be devoted to a case study of hierarchical modeling, the workhorse of applied Bayesian statistics. We will show how hierarchical models pool estimates toward the population means based on population variance and how this automatically estimates regularization and adjusts for multiple comparisons. The focus will be on probabilistic inference, and in particular on testing posterior predictive calibration and the sharpness of predictions.

Installing RStan: Please show up with RStan installed. Instructions are linked from here:

Warning: follow the instructions step-by-step; even though installation involves a CRAN package, it’s more complex than just installing from RStudio because a C++ toolchain is required at runtime.

If you run into trouble, please ask for help on our forums—they’re very friendly:

Full Day Schedule

10:00-11:00am Open Seminar – Introduction to the “Stan” System

11:00-11:30am Break

11:30am-1:00pm Tutorial part 1

1:00pm -2:00pm Lunch Break

2:00pm -3:30pm Tutorial part 2

3:30pm -3:45pm Break

3:45pm -5:30pm Tutorial part 3

The post Halifax, NS, Stan talk and course Thu 19 Oct appeared first on Statistical Modeling, Causal Inference, and Social Science.

To leave a comment for the author, please follow the link and comment on their blog: R – Statistical Modeling, Causal Inference, and Social Science.

R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)