Pragmatic Bayes

Follow-up

• You can find the slides of the talk here.

Type: SeminarSpeaker: Martin ModrákDate & time: Tue May 13 2025, 13:00-14:30 (CEST)Place: Atrium, Nobels väg 12BURL: https://www.martinmodrak.cz/

This is an in-person event only. No zoom link will be provided.

In his talk, Martin will argue that the tools of Bayesian statistics are useful and worth of your attention, even if you are not willing to subscribe to the Bayesian interpretation of probability or other philosophical commitments commonly associated with The Bayesian Way™.

Matrin Modrák is a lecturer and statistician at the Second Faculty of Medicine at Charles University in Prague.

Why attend?

Instead of getting tied up in philosophical debates, Martin’s talk will focus on pragmatic benefits of modern Bayesian tools, including the ability to

• build complex models matching the structure of the data and domain knowledge
• ask questions that are impossible or very difficult to answer with frequentist tools
• provide exact guarantees on inferences even for small samples (i.e. when asymptotic results for large samples do not hold)
• subject our modelling assumptions to severe tests

We will also touch on situations that are commonly considered to be theoretically problematic for Bayesian inference (selective data reporting, optional early stopping) and discuss how much those are problems for a pragmatic use of Bayesian methods.