#29 Model Assessment, Non-Parametric Models, And Much More, with Aki Vehtari

Learning Bayesian Statistics

I’ll be honest here: I had a hard time summarizing this episode for you, and, let’s face it, it’s all my guest’s fault! Why? Because Aki Vehtari works on so many interesting projects that it’s hard to sum them all up, even more so because he was very generous with his time for this episode! But let’s try anyway, shall we?

So, Aki is an Associate professor in computational probabilistic modeling at Aalto University, Finland. You already heard his delightful Finnish accent on episode 20, with Andrew Gelman and Jennifer Hill, talking about their latest book, « Regression and other stories ». He is also a co-author of the popular and awarded book « Bayesian Data Analysis », Third Edition, and a core-developer of the seminal probabilistic programming framework Stan.

An enthusiast of open-source software, Aki is a core-contributor to the ArviZ package and has been involved in many free software projects such as GPstuff for Gaussian processes and ELFI for likelihood inference.

His numerous research interests are Bayesian probability theory and methodology, especially model assessment and selection, non-parametric models (such as Gaussian processes), feature selection, dynamic models, and hierarchical models.

We talked about all that — and more — on this episode, in the context of his teaching at Aalto and the software-assisted Bayesian workflow he’s currently working on with a group of researchers.

Our theme music is « Good Bayesian », by Baba Brinkman (feat MC Lars and Mega Ran). Check out his awesome work at https://bababrinkman.com/ !

Thank you to my Patrons for making this episode possible!

Yusuke Saito, Avi Bryant, Ero Carrera, Brian Huey, Giuliano Cruz, Tim Gasser, James Wade, Tradd Salvo, Adam Bartonicek, William Benton, Alan O'Donnell, Mark Ormsby, Demetri Pananos, James Ahloy, Jon Berezowski, Robin Taylor, Thomas Wiecki, Chad Scherrer, Vincent Arel-Bundock, Nathaniel Neitzke, Zwelithini Tunyiswa, Elea McDonnell Feit, Bertrand Wilden, James Thompson, Stephen Oates, Gian Luca Di Tanna, Jack Wells, Matthew Maldonado, Ian Costley, Ally Salim, Larry Gill, Joshua Duncan, Ian Moran, Paul Oreto, Colin Caprani, George Ho and Colin Carroll.

Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)

Links from the show:

• New podcast website: https://www.learnbayesstats.com/
• Rate LBS on Podchaser: https://www.podchaser.com/podcasts/learning-bayesian-statistics-932588
• Aki's website: https://users.aalto.fi/~ave/
• Aki's educational material: https://avehtari.github.io/
• Aki on GitHub: https://github.com/avehtari
• Aki on Twitter: https://twitter.com/avehtari
• Bayesian Data Analysis, 3rd edition: https://www.routledge.com/Bayesian-Data-Analysis/Gelman-Carlin-Stern-Dunson-Vehtari-Rubin/p/book/9781439840955
• Bayesian Data Analysis course material: https://github.com/avehtari/BDA_course_Aalto
• Regression and Other Stories: https://avehtari.github.io/ROS-Examples/
• Aki’s favorite scientific books (so far): https://statmodeling.stat.columbia.edu/2018/05/14/aki_books/
• Aki's talk on Agile Probabilistic Programming: https://www.youtube.com/watch?v=cHlPgHn6btg
• Aki's posts on Andrew Gelman's blog: https://statmodeling.stat.columbia.edu/author/aki/
• Stan software: https://mc-stan.org/
• GPstuff - Gaussian process models for Bayesian analysis: https://research.cs.aalto.fi/pml/software/gpstuff/
• Projpred -- R package to perform projection predictive variable selection for GLMs: https://github.com/stan-dev/projpred