Measuring the quality of MCMC output


Markov chain Monte Carlo (MCMC) methods are now employed in nearly all fields of science. Along with a constant development of new methods, research on MCMC has led to steady improvements in our theoretical understanding of these techniques. However, most theoretical tools translate only vaguely into guidelines for practitioners, for example regarding the tuning of parameters and the choice of iterations to perform. How to address questions of “quality control” for practical MCMC algorithms has seemed to attract less attention than the development of new algorithms.

This event puts a spotlight on the question of assessing the quality of MCMC samples. It brings together 1) people who work on the development of methods addressing that question, and 2) people who encounter challenging sampling problems in applications. The goal is to bridge methodology and practice, so that methodologists can update their benchmarks and practitioners can update their toolbox. The workshop is intended to highlight advances, put people in touch, share experiences, identify challenges, and start collaborations among participants.

Dates, format, instructions

The workshop will be fully online. This website will collect relevant material, including posters submitted by participants and a link to a YouTube channel for pre-recorded video presentations. This content will be freely available and we hope that this website will remain a useful portal to this topic long after the workshop!

You can submit a (~20 minute) video recording of a research presentation and/or a poster by emailing the organisers (see addresses below). For videos, these materials should preferably be an mp4 file or a YouTube link, in addition to a PDF file for slides. Poster submissions should be PDF files.

All participants are encouraged to share some material, in order to facilitate interactions during the live sessions. If you are too shy to send a video or a poster, consider sharing some slides describing your research interests in connection to the workshop topic. You can also send us a link to a paper or to some code, and we will paste it here.

The workshop event itself will be made of three live sessions on Zoom and, on different dates and times so as to accommodate different time zones. Each session will each have a different theme, and will be organized roughly as:

  • introductory words (a bit longer during the first session),
  • a bunch of very short presentations advertising prerecorded talks, posters, papers etc,
  • a discussion, structured along questions specific to each session's theme,
  • a poster session on (open for at least 90 minutes each time)

The live sessions will be as follows.

  • Session 1: 11am-2pm UTC on Wednesday 6th October.
    Theme: general view on MCMC diagnostics. The first session will be an opportunity to comment on the meaning of "convergence diagnostics", review old and new methods and their limitations.
  • Session 2: 1pm-4pm UTC on Thursday 14th October.
    Theme: diagnostics for specific samplers. Here the focus will be on modern samplers, and the specific challenges that they raise when it comes to convergence diagnostics. For example: MCMC for phylogenetic inference, piecewise-deterministic Markov processes, gradient-based MCMC.
  • Session 3: 3pm-6pm UTC on Friday 22nd October.
    Theme: Theory and practice, bridging gaps, and ways forward. The last session will be an opportunity to review emerging practical challenges and discuss ways forward for practitioners, methodologists, and theoreticians, identifying fruitful avenues of research, collaborations.

There will be a prize for best presentation or poster, awarded to a junior researcher, selected by the advisors of the workshop listed below! The winner (or winners) will be announced shortly after the end of the workshop.

Youtube channel

We have set up a youtube playlist as part of the ISBA Youtube account , with the shared pre-recorded videos (in no particular order). We'd be happy to add your video! And thanks ISBA!

Youtube playlist for the workshop

Confirmed participants

The following individuals are confirmed participants and advisors for the workshop.

The following individuals are confirmed participants of the workshop.


Registration is free but compulsory, so as to limit the number of participants to an approximate maximum of 100 (as of October 2nd we count 76 registrations). We are primarily welcoming participants whose current research aligns with the topic of the workshop.

Please fill out the registration form here .

Organizers and Sponsors

The event is primarily organized by Leah South, Queensland University of Technology (QUT), and Pierre E. Jacob, ESSEC Business School, with the support of the BayesComp section of ISBA. We can be contacted at l1.south (at) and (at)

Free registration is courtesy of the QUT Centre for Data Science.