If you’re in NYC or Sidney, there are some Stan-related talks in the next few weeks.
New York
- 25 February. Jonah Gabry: shinyStan: a graphical user interface for exploring Bayesian models after MCMC. Register Now: New York Open Statistical Programming Meetup.
- 12 March. Rob Trangucci: #5: Non-centered parameterization aka the “Matt trick.” Register Now: Stan Users NYC Meetup.
Sydney
- 4 March. Bob Carpenter: The Benefits of a Probabilistic Model of Data Annotation. Macquarie Uni Computer Science Dept.
- 10 March, 2–3 PM. Bob Carpenter: Stan: Bayesian Inference Made Easy. Macquarie Uni Statistics Dept. Building E4A, room 523.
- 11 March, 6 PM, Mithcell Theatre, Level 1 at SMSA (Sydney Mechanics’ School of Arts). Bob Carpenter: RStan: Bayesian Inference Made Easy. Register Now: Sydney Users of R Forum (SURF) (Meetup)
Daniel,
WIll any of these be recorded and made available online? – I’d certainly be interested in seeing the Matt trick talk.
I don’t have any plans on recording Rob’s talk. Jonah’s talk last night was great and we’ll post a link if it was recorded.
I don’t think any of mine will be recorded, though NICTA says they’ll release the summer school videos, which is 2.5 hours of me on Bayes and MCMC, followed by 1.5 hours on Stan, followed by a fumbling 1 hour live demo of Stan.
As to the “Matt trick” talk, we call it the non-centered parameterization now, following previous literature (for once, Matt didn’t get there first!). But you’re in luck, because Michael Betancourt wrote it up as a paper, which is available on arXiv: http://arxiv.org/pdf/1312.0906.pdf
Sydney?
Yes, Sydney, Australia. I fixed the typo. I’m keeping my full schedule up to date on the post Stan Down Under.