Our meetings will be held tentatively every three weeks on Tuesdays 9.00-10.15 (during September-October, and weekday will change after October).

The idea is to meet first thing in the morning before heading to work. The location alternates between the department of Computer Science and Engineering (Lindholmen campus) and the department of Mathematical Sciences (Johanneberg campus).

Everyone is welcome to attend

*(you do not need to have read the paper in order to participate, though that's useful!)**.*Notice, topics are chosen according to the interests of the organizers, and it is assumed*some*knowledge of Bayesian statistics and computational (Monte Carlo) methods.**Subscribe**to our

**mailing list**to keep up to date! (

**any**mail address can be provided, not necessarily gmail).

**Interested in giving a talk on a notable paper?**Send your suggestions for consideration to

*picchini (at) chalmers.se*(maths dept) or

*richard.torkar (at) cse.gu.se*(CSE dept)

**Forthcoming events:**

**Past events:**

- Date: 1 October 2019, 9.00-10.15

Paper: Vehtari et al. "Rank normalization, folding and localization: an improved \hat{R} for assessing convergence of MCMC" https://arxiv.org/abs/1903.08008

Presented by: Richard Torkar (division Software Engineering, Chalmers/GU)

Location: room MVL14 (ground floor),

**Mathematical Sciences**, Chalmers tvärgata 3 https://goo.gl/maps/sprtwKL2o5UHwLSw6

- Date: 10 September 2019, 9.00-10.15

Paper: Fasiolo et al. "A comparison of inferential methods for highly nonlinear state-space models". https://arxiv.org/abs/1411.4564

Why this paper: it is a nice, approachable review on some inference methods for stochastic dynamical models (state-space models here). Some of the considered methods work also for non-state space models. It is a good opportunity to look at inferential possibilities based on plug-and-play approaches that are becoming more and more standard.

Why this paper: it is a nice, approachable review on some inference methods for stochastic dynamical models (state-space models here). Some of the considered methods work also for non-state space models. It is a good opportunity to look at inferential possibilities based on plug-and-play approaches that are becoming more and more standard.

Presented by: Umberto Picchini (dept. mathematics, Chalmers/GU)

Location: room 473, Jupiter building (4th floor),

**Computer Science and Engineering**, in Teknikgatan 7 https://goo.gl/maps/Yp6Ks3PXUh5WhyRd8