[Forum SIS] seminar: "Scalable Bayesian inference for dynamic state-space mixed-effects models", Sebastian Persson, 30 November @14:00

[Umberto Picchini]

You are welcome to the next Statistics seminar at Dept of Mathematical 
Sciences at Chalmers and Göteborg University.

We are very glad to have, on Tuesday 30 November,Sebastian Persson 
(Chalmers and University of Göteborg) who will talk of:

/"//Scalable Bayesian inference for dynamic state-space mixed-effects 
models//"/

*S. PerssonAbout the speaker: *Sebastian Persson 
<https://www.chalmers.se/en/staff/Pages/sebpe.aspx>'s research is about 
using mathematical modelling for understanding different cellular 
processes that are connected to ageing in yeast. For example, using 
reaction-diffusion models to study cell polarization, or ODE- and 
SDE-modelling and nonlinear mixed-effects modelling to study signaling 
pathways that are connected to glucose sensing.

*Abstract:* Parameter inference is an important step when constructing a 
dynamic model in many fields ranging from biology, medicine (PK/PD) to 
finance. In many scenarios, such as when modelling biological 
single-cell behaviour, we are interested in inference for an entire 
population by simultaneously fitting observations from multiple 
individuals. However, inference from multi-individual longitudal data is 
often non-trivial due to the presence of intrinsic and extrinsic sources 
of variability. To address this, we consider inference for the 
challenging case when the dynamics for a state-space mixed-effects model 
(SSMEM) are driven by stochastic processes such as a Markov-jump process 
or a stochastic differential equation. We present an efficient 
Gibbs-sampler for fully Bayesian inference for SSMEMs, which compared to 
previous samplers can be more than 30 times faster for many (>100) 
individuals. The individual parameters in the Gibbs-sampler, which have 
an intractable likelihood, are efficiently sampled via 
correlated-particle pseudo-marginal Metropolis-Hastings' steps. The 
population parameters of the random effects, which have a tractable 
likelihood, are updated using a HMC sampler to allow for a realistic 
parameterization of the individual parameters.

The performance of our Gibbs sampler is investigated on challenging 
simulated datasets (e.g., a stochastic bi-stable model) and on a 
real-life dataset. Furthermore, we investigate the performance of 
different adaptive MCMC algorithms for the pseudo-marginal steps.
*
*
*
*
*Feel free to spread this announcement in your network.*
*
*
*Where*: room MVL15 or https://chalmers.zoom.us/j/64449172943
  Password: 770823
**

****
*When*: Tuesday 30 November at 14.00-15.00 (Swedish time).
**
**

-- 
_________________________________________________________
Umberto Picchini, Associate Professor, PhD, Docent
https://umbertopicchini.github.io/  , twitter: @uPicchini
-------------- parte successiva --------------
Un allegato HTML è stato rimosso...
URL: <http://www.stat.unipg.it/pipermail/sis/attachments/20211127/790e3beb/attachment.html>
-------------- parte successiva --------------
Un allegato non testuale è stato rimosso....
Nome:        sebastianpersson.jpg
Tipo:        image/jpeg
Dimensione:  80507 bytes
Descrizione: non disponibile
URL:         <http://www.stat.unipg.it/pipermail/sis/attachments/20211127/790e3beb/attachment.jpg>