Modeling and Simulation Group Meeting Old

MCMC in infinite dimensions and data visualization

Speaker: Sonia Reilly and Mathias Kasiulis

Location: Warren Weaver Hall 1302

Date: Thursday, April 27, 2023, 12:45 p.m.


Sonia Reilly will talk about Ensemble MCMC sampling for infinite-dimensional Bayesian inverse problem. Sonis will describe the problem of sampling posterior distributions from infinite-dimensional Bayesian inverse problems and propose a solution that uses an ensemble of MCMC chains to improve on existing MCMC methods.

Mathias Kasiulis will tell us about Some simple ideas on data visualization and scientific images.
As scientists, one of our key duties is sharing our results in a clear and pedagogical way to audiences, often via visual documents like articles or slides. As a result, our choices of data visualization can end up being the most important factor to determine whether our work actually reaches a broad, or even any, audience. In this short tutorial talk, Mathias will give a few leads on how to present data in a clear, appealing way -- and how it might even change the way you look at your own research.