RStan: Statistical Modeling Made Easy with Bob Carpenter

This August we'll be joined by special guest Dr. Bob Carpenter to learn about Stan (http://mc-stan.org/), his open-source probabilistic programming language, and how to use it with R.

RStan: Statistical Modeling Made Easy Bob Carpenter
Columbia Uni., Dept. of Statistics

I'll introduce Stan, a new language for expressing statistical
models with support for full Bayesian inference via sampling (Hamiltonian Monte Carlo) and maximum likelihood estimation via optimization (L-BFGS) and curvature (higher-order autodiff). I'll begin with an introduction to Bayesian modeling and how it preserves estimation uncertainty in posterior inference, focusing on event probability estimation, decision theory, and out-of-sample prediction. I'll introduce Stan's modeling language based on examples, working up from the very basics (Bernoulli model of binary outcomes, simple linear regression), through our bread and butter algorithms (multilevel generalized linear models, aka mixed effects), and finishing with some advanced model examples (Gaussian processes, diff eq physical system models). I'll concentrate on how Stan can be invoked through R using the RStan interface. I'll also provide a high-level overview of the back end, touching on how models get translated to C++, how automatic differentiation allows us to compute gradients up to machine precision, how variables are transformed to unconstrained support on the real numbers, and how our adaptive Hamiltonian Monte Carlo works using animations. Time permitting, I'll introduce
ShinyStan, our interactive posterior visualization tool, and describe the latest edition to Stan, black-box variational inference. I'll wrap up with an overview of what we have in store over the next couple of years.

About Bob
Bob Carpenter is a research scientist in computational statistics (Columbia University). He designed the Stan probabilistic programming language and is one of the Stan core developers.

Bob has a Ph.D. in cognitive and computer science (University of Edinburgh), worked as a professor of computational linguistics (Carnegie Mellon University), an industrial researcher and programmer in speech recognition and natural language processing (Bell Labs,
SpeechWorks, LingPipe).

We'll start the night off with some announcements, and leave the rest of the time for Bob's presentation followed by Q&A.

Join us next door at Hopcat for food & drinks after the meetup!

Directions/Parking

The entrance to Barracuda is located on the East side of Maynard St., underneath the parking deck between Hopcat and Nickel's Arcade. The outer doors should be unlocked, but someone will have to let you in the inner doors; we'll make sure someone is stationed there until sometime after 7pm.

Street parking is free after 6pm, but can be hard to find in this area due to all the restaurants nearby. The easiest parking is the Maynard/Thompson parking structure, accessible from Maynard or Thompson. If you park there and exit on Maynard you simply walk across the street to get here. If that is structure is full or inconvenient, there are also structures on Washington St (between State and Fifth) and Division/Fifth (enter from either, between Liberty and Williams).