#87 Unlocking the Power of Bayesian Causal Inference, with Ben Vincent
Learning Bayesian Statistics
The Role of Causal Inference in Understanding Statistical Paradoxes and High-Stakes Interventions
Models can be described as directed acyclic graphs in Bayesian inference./nUnderstanding causal thinking can provide intuitive explanations for statistical paradoxes./nCausal inference is most helpful when trying to gain a better understanding of data./nCausal inference is especially valuable in high stakes interventions.
Proudly sponsored by PyMC Labs, the Bayesian Consultancy. Book a call, or get in touch!
I’ll be honest — this episode is long overdue. Not only because Ben Vincent is a friend, fellow PyMC Labs developer, and outstanding Bayesian modeler. But because he works on so many fascinating topics — so I’m all the happier to finally have him on the show!
In this episode, we’re gonna focus on causal inference, how it naturally extends Bayesian modeling, and how you can use the CausalPy open-source package to supercharge your Bayesian causal inference. We’ll also touch on marketing models and the pymc-marketing package, because, well, Ben does a lot of stuff ;)
Ben got his PhD in neuroscience at Sussex University, in the UK. After a postdoc at the University of Bristol, working on robots and active vision, as well as 15 years as a lecturer at the Scottish University of Dundee, he switched to the private sector, working with us full time at PyMC Labs — and that is a treat!
When he’s not working, Ben loves running 5k’s, cycling in the forest, lifting weights, and… learning about modern monetary theory.
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, Giuliano Cruz, Tim Gasser, James Wade, Tradd Salvo, William Benton, James Ahloy, Robin Taylor,, Chad Scherrer, Nathaniel Neitzke, Zwelithini Tunyiswa, Bertrand Wilden, James Thompson, Stephen Oates, Gian Luca Di Tanna, Jack Wells, Matthew Maldonado, Ian Costley, Ally Salim, Larry Gill, Ian Moran, Paul Oreto, Colin Caprani, Colin Carroll, Nathaniel Burbank, Michael Osthege, Rémi Louf, Clive Edelsten, Henri Wallen, Hugo Botha, Vinh Nguyen, Raul Maldonado, Marcin Elantkowski, Adam C. Smith, Will Kurt, Andrew Moskowitz, Hector Munoz, Marco Gorelli, Simon Kessell, Bradley Rode, Patrick Kelley, Rick Anderson, Casper de Bruin, Philippe Labonde, Michael Hankin, Cameron Smith, Tomáš Frýda, Ryan Wesslen, Andreas Netti, Riley King, Yoshiyuki Hamajima, Sven De Maeyer, Michael DeCrescenzo, Fergal M, Mason Yahr, Naoya Kanai, Steven Rowland, Aubrey Clayton, Jeannine Sue, Omri Har Shemesh, Scott Anthony Robson, Robert Yolken, Or Duek, Pavel Dusek, Paul Cox, Trey Causey, Andreas Kröpelin, Raphaël R, Nicolas Rode, Gabriel Stechschulte, Arkady, Kurt TeKolste, Gergely Juhasz, Marcus Nölke, Maggi Mackintosh, Grant Pezzolesi, Avram Aelony and Joshua Meehl.
Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)
Links from the show:
- Ben’s website: https://drbenvincent.github.io/
- Ben on GitHub: https://github.com/drbenvincent
- Ben on Twitter: https://twitter.com/inferencelab
- Ben on LinkedIn: https://www.linkedin.com/in/dr-benjamin-vincent-503571127/
- CausalPy – Causal inference for quasi-experiments: https://causalpy.readthedocs.io/en/latest/
- PyMC Marketing – Bayesian marketing toolbox in PyMC: https://www.pymc-marketing.io/en/stable/index.html
- PyMC Labs : https://www.pymc-labs.io/products/
- LBS #23 – Bayesian Stats in Business and Marketing Analytics, with Elea McDonnel Feit: https://learnbayesstats.com/episode/23-bayesian-stats-in-business-and-marketing-analytics-with-elea-mcdonnel-feit/
- LBS #63 – Media Mix Models & Bayes for Marketing, with Luciano Paz: https://learnbayesstats.com/episode/63-media-mix-models-bayes-marketing-luciano-paz/
Abstract
written by Christoph Bamberg
In this podcast episode, our guest, Ben Vincent, a fellow member of PyMC Labs with a PhD in Neuroscience and extensive experience in teaching and data analysis of course, introduces us to CausalPy and PyMC Marketing.
During his academic career, Ben got introduced to Bayesian statistics but, like most academics, did not come across causal inference.
We discuss the importance of a systematic causal approach for important questions like health care interventions or marketing investments.
Although causality is somewhat orthogonal to the choice of statistical approach, Bayesian statistics is a good basis for causal analyses, for example in the for of Directed Acyclical Graphs.
To make causal inference more accessible, Ben developed a Python package called CausalPy, which allows you perform common causal inferences, e.g. working with natural experiments.
Ben was also involved in the development of PyMC Marketing, a package that conveniently bundles important analysis capacities for Marketing. The package focuses on Media Mix Modelling and customer lifetime analysis.
We also talked about his extensive experience teaching statistics at university and current teaching of Bayesian methods in industry. His advice to students is to really engage with your learning material, coding through examples, making the learning more pleasurable and practical.
Transcript
Please note that this is an automated transcript that may contain errors. Feel free to reach out if you're willing to correct them.
Remember Everything You Learn from Podcasts
