Learning Bayesian Statistics

Episode sponsored by Paperpile: paperpile.comGet 20% off until December 31st with promo code GOODBAYESIAN21I don’t know if you’ve heard, but there is a virus that took over most of the world in the past year? I haven’t dedicated any episode to Covid yet. First because research was moving a lot — and fast. And second because modeling Covid is very, very hard.But we know more about it now, so I thought it was a good time to pause and ponder — how does the virus circulate? How can we model it and, ultimately, defeat it? What are the challenges in doing so?To talk about that, I had the chance to host Michael Osthege and Thomas Vladeck, who both were part of the team who developed the Rt-live model, a Bayesian model to infer the reproductive rate of Covid19 in the general population. As you’ll hear, modeling the evolution of this virus is challenging, fascinating, and a perfect fit for Bayesian modeling! It truly is a wonderful example of Bayesian generative modeling.Tom is the Managing Director of Gradient Metrics, a quantitative market research firm, and a Co-Founder of Recast, a media mix model for modern brands.Michael is a PhD student in laboratory automation and bioprocess optimization at the Forschungszentrum Jülich in Germany, and a fellow PyMC core-developer. As he works a lot on the coming brand new version 4, we’ll take this opportunity to talk about the current developments and where the project is headed.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, 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, Colin Carroll, Nathaniel Burbank, Michael Osthege, Rémi Louf, Clive Edelsten, Henri Wallen, Jonathan Sedar, Hugo Botha, Vinh Nguyen, Raul Maldonado, Marcin Elantkowski, Tim Radtke, Adam C. Smith, Will Kurt, Andrew Moskowitz, Hector Munoz, Marco Gorelli, Simon Kessell, Bradley Rode and Patrick Kelley.Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)Links from the show:Tom on Twitter: https://twitter.com/tvladeckTom's newsletter: https://tvladeck.substack.com/Michael on Twitter: https://twitter.com/theCakeMichael on GitHub: https://github.com/michaelosthegeRt Live dashboard: https://rtlive.de/global.htmlRt Live model tutorial: https://github.com/rtcovidlive/rtlive-global/blob/master/notebooks/Tutorial_model.ipynbRt Live model code: https://github.com/rtcovidlive/rtlive-globalEstimating Rt: https://staff.math.su.se/hoehle/blog/2020/04/15/effectiveR0.htmlGreat resource on terminology: https://royalsociety.org/-/media/policy/projects/set-c/set-covid-19-R-estimates.pdf?la=en-GB&hash=FDFFC11968E5D247D8FF641930680BD6Using Hierarchical Multinomial Regression to Predict Elections in Paris districts: https://www.youtube.com/watch?v=EYdIzSYwbSwLBS #34, Multilevel Regression, Post-stratification & Missing Data, with Lauren Kennedy: https://www.learnbayesstats.com/episode/34-multilevel-regression-post-stratification-missing-data-lauren-kennedymrmp - Multilevel Regression and Marginal Poststratification: https://rdrr.io/github/jwyatt85/MRmP/man/mrmp.htmlAutomating daily runs for rt.live’s COVID-19 data using Airflow & ECS: https://medium.com/@mikekrieger/automating-daily-runs-for-rt-lives-covid-19-data-dcda26ed2e2eLBS #23, Bayesian Stats in Business and Marketing Analytics, with Elea McDonnel Feit: https://www.learnbayesstats.com/episode/23-bayesian-stats-in-business-and-marketing-analytics-with-elea-mcdonnel-feit

Episode sponsored by Paperpile: paperpile.comGet 20% off until December 31st with promo code GOODBAYESIAN21We often talk about applying Bayesian statistics on this podcast. But how do we teach them? What’s the best way to introduce them from a young age and make sure the skills students learn in the stats class are transferable?Well, lucky us, Mine Dogucu’s research tackles precisely those topics!An Assistant Professor of Teaching in the Department of Statistics at University of California Irvine, Mine is both an educator with an interest in statistics, and an applied statistician with experience in educational research.Her work focuses on modern pedagogical approaches in the statistics curriculum, making data science education more accessible. In particular, she teaches an undergraduate Bayesian course, and is the coauthor of the upcoming book Bayes Rules! An Introduction to Bayesian Modeling with R.In other words, Mine is not only interested in teaching, but also in how best to teach statistics – how to engage students in remote classes, how to get to know them, how to best record and edit remote courses, etc. She writes about these topics on her blog, DataPedagogy.com.She also works on accessibility and inclusion, as well as a study that investigates how popular Bayesian courses are at the undergraduate level in the US — that should be fun to talk about!Mine did her Master’s at Bogazici University in Istanbul, Turkey, and then her PhD in Quantitative Research, Evaluation, and Measurement at Ohio State University.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, 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, Colin Carroll, Nathaniel Burbank, Michael Osthege, Rémi Louf, Clive Edelsten, Henri Wallen, Jonathan Sedar, Hugo Botha, Vinh Nguyen, Raul Maldonado, Marcin Elantkowski, Tim Radtke, Adam C. Smith, Will Kurt, Andrew Moskowitz, John Johnson, Hector Munoz, Marco Gorelli, Simon Kessell, Bradley Rode and Patrick Kelley.Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)Links from the show:Mine's website: https://mdogucu.ics.uci.edu/index.htmlMine's blog: https://www.datapedagogy.com/Mine on Twitter: https://twitter.com/MineDogucuMine on GitHub: https://github.com/mdogucuBayes Rules! An Introduction to Bayesian Modeling with R: https://www.bayesrulesbook.com/R package for Supplemental Materials for the Bayes Rules! Book: https://github.com/bayes-rules/bayesrulesStats 115 - Introduction to Bayesian Data Analysis: https://www.stats115.com/Undergraduate Bayesian Education Network: https://undergrad-bayes.netlify.app/network.htmlWorkshop "Teaching Bayesian Statistics at the Undergraduate Level": https://www.causeweb.org/cause/uscots/uscots21/workshop/4

Let’s think Bayes, shall we? And who better to do that than the author of the well known book, Think Bayes — Allen Downey himself! Since the second edition was just released, the timing couldn’t be better!Allen is a professor at Olin College and the author of books related to software and data science, including Think Python, Think Bayes, and Think Complexity. His blog, Probably Overthinking It, features articles on Bayesian probability and statistics. He holds a Ph.D. from U.C. Berkeley, and bachelors and masters degrees from MIT.In this special episode, Allen and I talked about his background, how he came to the stats and teaching worlds, and why he wanted to write this book in the first place. He’ll tell us who this book is written for, what’s new in the second edition, and which mistakes his students most commonly make when starting to learn Bayesian stats. We also talked about some types of models, their usefulness and their weaknesses, but I’ll let you discover that.Now for another good news: 5 Patrons of the show will get Think Bayes for free! To qualify, you just need to go the form I linked to in the 'Learn Bayes Stats' Slack channel or the Patreon page and enter your email address. That’s it. After a week or so, Allen and I will choose 5 winners at random, who will receive the book for free!If you’re not a Patron yet, make sure to check out patreon.com/learnbayesstats if you don’t want to miss out on these goodies!And even if you’re not a Patron, I love you dear listeners, so you all get a discount when you go buy the book at https://www.learnbayesstats.com/buy-think-bayes (unfortunately, this only applies for purchases in the US and Canada).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, 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, Colin Carroll, Nathaniel Burbank, Michael Osthege, Rémi Louf, Clive Edelsten, Henri Wallen, Jonathan Sedar, Hugo Botha, Vinh Nguyen, Raul Maldonado, Marcin Elantkowski, Tim Radtke, Adam C. Smith, Will Kurt, Andrew Moskowitz, John Johnson and Hector Munoz.Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)Links from the show:Give LBS a 5-star rating on Podchaser: https://www.podchaser.com/learnbayesstatsBuy Think Bayes at a 40% discount with the code LBS40 (expires on July 31; only applies for purchases in the US and Canada): https://www.learnbayesstats.com/buy-think-bayesThink Bayes 2 online: http://allendowney.github.io/ThinkBayes2/index.htmlAllen's blog: https://www.allendowney.com/blog/Allen on Twitter: https://twitter.com/allendowneyAllen on GitHub: https://github.com/AllenDowneyInformation theory, inference and learning algorithms, David MacKay: https://www.inference.org.uk/itila/Statistical Rethinking, Richard McElreath: http://xcelab.net/rm/statistical-rethinking/Doing Bayesian Data Analysis, John Kruschke: https://sites.google.com/site/doingbayesiandataanalysis/homeProbabilistic Programming & Bayesian Methods for Hackers, Cam Davidson-Pilon: http://camdavidsonpilon.github.io/Probabilistic-Programming-and-Bayesian-Methods-for-Hackers/LBS #14, Hidden Markov Models & Statistical Ecology, with Vianey Leos-Barajas: https://www.learnbayesstats.com/episode/14-hidden-markov-models-statistical-ecology-with-vianey-leos-barajasThe Prosecutor's fallacy: https://en.wikipedia.org/wiki/Prosecutor%27s_fallacyConfidence intervals vs. Bayesian intervals, E.T. Jaynes: https://bayes.wustl.edu/etj/articles/confidence.pdfSuperforecasting, The Art and Science of Prediction, Philip Tetlock: https://en.wikipedia.org/wiki/Superforecasting:_The_Art_and_Science_of_Prediction

We all know about these accidental discoveries — penicillin, the heating power of microwaves, or the famous (and delicious) tarte tatin. I don’t know why, but I just love serendipity. And, as you’ll hear, this episode is deliciously full of it…Thanks to Allison Hilger and Timo Roettger, we’ll discover the world of linguistics, how Bayesian stats are helpful there, and how Paul Bürkner’s BRMS package has been instrumental in this field. To my surprise — and perhaps yours — the speech and language sciences are pretty quantitative and computational!As she recently discovered Bayesian stats, Allison will also tell us about the challenges she’s faced from advisors and reviewers during her PhD at Northwestern University, and the advice she’d have for people in the same situation.Allison is now an Assistant Professor at the University of Colorado Boulder. The overall goal in her research is to improve our understanding of motor speech control processes, in order to inform effective speech therapy treatments for improved speech naturalness and intelligibility. Allison also worked clinically as a speech-language pathologist in Chicago for a year. As a new Colorado resident, her new hobbies include hiking, skiing, and biking — and then reading or going to dog parks when she’s to tired.Holding a PhD in linguistics from the University of Cologne, Germany, Timo is an Associate Professor for linguistics at the University of Oslo, Norway. Timo tries to understand how people communicate their intentions using speech – how are speech signals retrieved; how do people learn and generalize? Timo is also committed to improving methodologies across the language sciences in light of the replication crisis, with a strong emphasis on open science.Most importantly, Timo loves hiking, watching movies or, even better, watching people play video games!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, 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, Colin Carroll, Nathaniel Burbank, Michael Osthege, Rémi Louf, Clive Edelsten, Henri Wallen, Jonathan Sedar, Hugo Botha, Vinh Nguyen, Raul Maldonado, Marcin Elantkowski, Tim Radtke, Adam C. Smith, Will Kurt and Andrew Moskowitz.Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)Links from the show:Allison's website: https://allisonhilger.com/Allison on Twitter: https://twitter.com/drahilgerAllison's motor speech lab: https://www.colorado.edu/lab/motor-speech/Timo's website: https://www.simplpoints.com/Timo on Twitter: https://twitter.com/TimoRoettgerBayesian regression modeling (for factorial designs) -- A tutorial: https://psyarxiv.com/cdxv3An Introduction to Bayesian Multilevel Models Using brms -- A Case Study of Gender Effects on Vowel Variability in Standard Indonesian: https://biblio.ugent.be/publication/8624552/file/8624553.pdfLongitudinal Growth in Intelligibility of Connected Speech From 2 to 8 Years in Children With Cerebral Palsy -- A Novel Bayesian Approach: https://pubs.asha.org/doi/10.1044/2020_JSLHR-20-00181LBS #35 The Past, Present & Future of BRMS, with Paul Bürkner: https://www.learnbayesstats.com/episode/35-past-present-future-brms-paul-burknerLBS #16 Bayesian Statistics the Fun Way, with Will Kurt: https://www.learnbayesstats.com/episode/16-bayesian-statistics-the-fun-way-with-will-kurtWill Kurt's Bayesian Statistics The Fun Way: https://nostarch.com/learnbayesLBS #20 Regression and Other Stories, with Andrew Gelman, Jennifer Hill & Aki Vehtari: https://www.learnbayesstats.com/episode/20-regression-and-other-stories-with-andrew-gelman-jennifer-hill-aki-vehtariRegression and Other Stories examples: https://avehtari.github.io/ROS-Examples/

Episode sponsored by Tidelift: tidelift.comIt’s been a while since we talked about biostatistics and bioinformatics on this podcast, so I thought it could be interesting to talk to Jacki Buros — and that was a very good idea!She’ll walk us through examples of Bayesian models she uses to, for instance, work on biomarker discovery for cancer immunotherapies. She’ll also introduce you to survival models — their usefulness, their powers and their challenges.Interestingly, all of this will highlight a handful of skills that Jacki would try to instill in her students if she had to teach Bayesian methods.The Head of Data and Analytics at Generable, a state-of-the-art Bayesian platform for oncology clinical trials, Jacki has been working in biostatistics and bioinformatics for over 15 years. She started in cardiology research at the TIMI Study Group at Harvard Medical School before working in Alzheimer’s Disease genetics at Boston University and in biomarker discovery for cancer immunotherapies at the Hammer Lab. Most recently she was the Lead Biostatistician at the Institute for Next Generation Health Care at Mount Sinai.An open-source enthusiast, Jacki is also a contributor to Stan and rstanarm, and the author of the survivalstan package, a library of Stan models for survival analysis.Last but not least, Jacki is an avid sailor and skier!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, 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, Colin Carroll, Nathaniel Burbank, Michael Osthege, Rémi Louf, Clive Edelsten, Henri Wallen, Jonathan Sedar, Hugo Botha, Vinh Nguyen, Raul Maldonado, Marcin Elantkowski, Tim Radtke, Adam C. Smith, Will Kurt and Andrew Moskowitz.Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)Links from the show:Nominate "Learn Bayes Stats" as "Best Podcast of 2021" and "Best Tech Podcast" by entering its Apple feed in this form!Jacki on Twitter: https://twitter.com/jackiburosJacki on GitHub: https://github.com/jburosJacki on Orcid: https://orcid.org/0000-0001-9588-4889survivalstan -- Survival Models in Stan: https://github.com/hammerlab/survivalstanrstanarm -- R model-fitting functions using Stan: http://mc-stan.org/rstanarm/Generable -- Bayesian platform for oncology clinical trials: https://www.generable.com/StanCon 2020 ArviZ presentation : https://github.com/arviz-devs/arviz_misc/tree/master/stancon_2020Thinking in Bets -- Making Smarter Decisions When You Don't Have All the Facts : https://www.goodreads.com/book/show/35957157-thinking-in-betsScott Kelly and his space travels (in French): https://www.franceculture.fr/emissions/la-methode-scientifique/la-methode-scientifique-mardi-30-janvier-2018Bayesian Workflow paper: https://arxiv.org/pdf/2011.01808v1.pdfBayesian Survival Analysis: https://www.springer.com/gp/book/9780387952772Bayesian Survival Analysis Using the rstanarm R Package: https://arxiv.org/pdf/2002.09633.pdfSurvival Analysis, A Self-Learning Text: https://www.springer.com/gp/book/9781441966452Survival and Event History Analysis, A Process Point of View: https://www.springer.com/gp/book/9780387202877Prognostic Significance of Tumor-Infiltrating B Cells and Plasma Cells in Human Cancer: https://clincancerres.aacrjournals.org/content/24/24/6125

Episode sponsored by Tidelift: tidelift.comImagine me rapping: "Let me show you how to be a good Bayesian. Change your predictions after taking information in, and if you’re thinking I’ll be less than amazing, let’s adjust those expectations!"What?? Nah, you’re right, I’m not as good as Baba Brinkman. Actually, the best to perform « Good Bayesian » live on the podcast would just be to invite him for an episode… Wait, isn’t that what I did???Well indeed! For this episode, I had the great pleasure of hosting rap artist, science communicator and revered author of « Good Bayesian », Baba Brinkman!We talked about his passion for oral poetry, his rap career, what being a good rapper means and the difficulties he encounters to establish himself as a proper rapper.Baba began his rap career in 1998, freestyling and writing songs in his hometown of Vancouver, Canada.In 2000 he started adapting Chaucer’s Canterbury Tales into original rap compositions, and in 2004 he premiered a one man show based on his Master’s thesis, The Rap Canterbury Tales, exploring parallels between hip-hop music and medieval poetry.Over the years, Baba went on to create “Rap Guides” dedicated to scientific topics, like evolution, consciousness, medicine, religion, and climate change – and I encourage you to give them all a listen!By the way, do you know the common point between rap and evolutionary biology? Well, you’ll have to tune in for the answer… And make sure you listen until the end: Baba has a very, very nice surprise for you!A little tip: if you wanna enjoy it to the fullest, I put the unedited video version of this interview in the show notes ;) By the way, let me know if you like these video live streams — I might just do them again if you do!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, Colin Carroll, Nathaniel Burbank, Michael Osthege, Rémi Louf, Clive Edelsten, Henri Wallen, Jonathan Sedar, Hugo Botha, Vinh Nguyen, Raul Maldonado, Marcin Elantkowski and Tim Radtke.Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)Links from the show:Video live-stream of the episode: https://www.youtube.com/watch?v=YkFXpP_SvHkBaba on Twitter: https://twitter.com/bababrinkmanBaba on YouTube: https://www.youtube.com/channel/UCz9Qm66ewnY0LAlZlL4HK9gBaba on Spotify: https://open.spotify.com/artist/7DqKchcLvOIgR87RzJm3XHBaba's website: https://bababrinkman.com/Event Rap Kickstarter: https://www.kickstarter.com/projects/bababrinkman/event-rap-the-one-stop-custom-rap-shopEvent Rap website: https://www.eventrap.com/Anil Seth -- Your Brain Hallucinates your Conscious Reality: https://www.ted.com/talks/anil_seth_your_brain_hallucinates_your_conscious_realityThe Big Picture -- On the Origins of Life, Meaning, and the Universe Itself: https://www.amazon.com/Big-Picture-Origins-Meaning-Universe/dp/1101984252

Episode sponsored by Tidelift: tidelift.comI don’t know about you, but the notion of time is really intriguing to me: it’s a purely artificial notion; we humans invented it — as an experiment, I asked my cat what time it was one day; needless to say it wasn’t very conclusive… And yet, the notion of time is so central to our lives — our work, leisures and projects depend on it.So much so that time series predictions represent a big part of the statistics and machine learning world. And to talk about all that, who better than a time master, namely Sean Taylor?Sean is a co-creator of the Prophet time series package, available in R and Python. He’s a social scientist and statistician specialized in methods for solving causal inference and business decision problems. Sean is particularly interested in building tools for practitioners working on real-world problems, and likes to hang out with people from many fields — computer scientists, economists, political scientists, statisticians, machine learning researchers, business school scholars — although I guess he does that remotely these days…Currently head of the Rideshare Labs team at Lyft, Sean was a research scientist and manager on Facebook’s Core Data Science Team and did a PhD in information systems at NYU’s Stern School of Business. He did his undergraduate at the University of Pennsylvania, studying economics, finance, and information systems. Last but not least, he grew up in Philadelphia, so, of course, he’s a huge Eagles fan! For my non US listeners, we’re talking about the football team here, not the bird!We also talked about two of my favorite topics — science communication and epistemology — so I had a lot of fun talking with Sean, and I hope you’ll deem this episode a good investment of your time 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, Colin Carroll, Nathaniel Burbank, Michael Osthege, Rémi Louf, Clive Edelsten, Henri Wallen, Jonathan Sedar, Hugo Botha, Vinh Nguyen and Raul Maldonado.Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)Links from the show:Sean's website: https://seanjtaylor.com/Sean on GitHub: https://github.com/seanjtaylorSean on Twitter: https://twitter.com/seanjtaylorProphet docs: https://facebook.github.io/prophet/Forecasting at Scale -- How and why we developed Prophet for forecasting at Facebook: https://www.youtube.com/watch?v=OaTAe4W9IfA Forecasting at Scale paper: https://www.tandfonline.com/doi/abs/10.1080/00031305.2017.1380080?journalCode=utas20&TimeSeers -- Hierarchical version of Prophet, written in PyMC3: https://github.com/MBrouns/timeseersThe Art of Doing Science and Engineering -- Learning to Learn: https://www.amazon.com/Art-Doing-Science-Engineering-Learning/dp/1732265178NeuralProphet -- Forecasting model based on Neural Networks in PyTorch: https://github.com/ourownstory/neural_prophet/Introducing PyMC Labs: https://www.pymc-labs.io/blog-posts/saving-the-world/

Episode sponsored by Tidelift: tidelift.comI bet you already heard of Bayesian nonparametric models, at least on this very podcast. We already talked about Dirichlet Processes with Karin Knudson on episode 4, and then about Gaussian Processes with Elizaveta Semenova on episode 21. Now we’re gonna dive into the mathematical properties of these objects, to understand them better — because, as you may know, Bayesian nonparametrics are quite powerful but also very hard to fit!Along the way, you’ll learn about probabilistic circuits, sum-product networks and — what a delight — you’ll hear from the Julia community! Indeed, my guest for this episode is no other than… Martin Trapp!Martin is a core developer of Turing.jl, an open-source framework for probabilistic programming in Julia, and a post-doc in probabilistic machine learning at Aalto University, Finland.Martin loves working on sum-product networks and Bayesian non-parametrics. And indeed, his research interests focus on probabilistic models that exploit structural properties to allow efficient and exact computations while maintaining the capability to model complex relationships in data. In other words, Martin’s research is focused on tractable probabilistic models.Martin did his MsC in computational intelligence at the Vienna University of Technology and just finished his PhD in machine learning at the Graz University of Technology. He doesn’t only like to study the tractability of probabilistic models — he also is very found of climbing!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, Colin Carroll, Nathaniel Burbank, Michael Osthege, Rémi Louf, Clive Edelsten, Henri Wallen, Jonathan Sedar, Hugo Botha, Vinh Nguyen and Raul Maldonado.Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)Links from the show:Martin's website: https://trappmartin.github.io/Martin on GitHub: https://github.com/trappmartinMartin on Twitter: https://twitter.com/martin_trappTuring, Bayesian inference with Julia: https://turing.ml/dev/Hierarchical Dirichlet Processes: https://people.eecs.berkeley.edu/~jordan/papers/hdp.pdfThe Automatic Statistician: https://www.doc.ic.ac.uk/~mpd37/teaching/2014/ml_tutorials/2014-01-29-slides_zoubin2.pdfTruncated Random Measures: https://arxiv.org/abs/1603.00861Deep Structured Mixtures of Gaussian Processes: https://arxiv.org/abs/1910.04536Probabilistic Circuits -- Representations, Inference, Learning and Theory: https://www.youtube.com/watch?v=2RAG5-L9R70Applied Stochastic Differential Equations, from Simo Särkkä and Arno Solin: https://users.aalto.fi/~asolin/sde-book/sde-book.pdf

Episode sponsored by Tidelift: tidelift.comOne of the most common guest suggestions that you dear listeners make is… inviting Paul Bürkner on the show! Why? Because he’s a member of the Stan development team and he created BRMS, a popular R package to make and sample from Bayesian regression models using Stan. And, as I like you, I did invite Paul on the show and, well, that was a good call: we had an amazing conversation, spanning so many topics that I can’t list them all here!I asked him why he created BRMS, in which fields it’s mostly used, what its weaknesses are, and which improvements to the package he’s currently working on. But that’s not it! Paul also gave his advice to people realizing that Bayesian methods would be useful to their research, but who fear facing challenges from advisors or reviewers.Besides being a Bayesian rockstar, Paul is a statistician working as an Independent Junior Research Group Leader at the Cluster of Excellence SimTech at the University of Stuttgart, Germany. Previously, he has studied Psychology and Mathematics at the Universities of Münster and Hagen and did his PhD in Münster about optimal design and Bayesian data analysis, and he also worked as a Postdoctoral researcher at the Department of Computer Science at Aalto University, Finland.So, of course, I asked him about the software-assisted Bayesian workflow that he’s currently working on with Aki Vehtari, which led us to no less than the future of probabilistic programming languages…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, Colin Carroll, Nathaniel Burbank, Michael Osthege, Rémi Louf, Clive Edelsten, Henri Wallen and Jonathan Sedar.Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)Links from the show:Paul's website: https://paul-buerkner.github.io/about/Paul on Twitter: https://twitter.com/paulbuerknerPaul on GitHub: https://github.com/paul-buerknerBRMS docs: https://paul-buerkner.github.io/brms/Stan docs: https://mc-stan.org/Bayesian workflow paper: https://arxiv.org/pdf/2011.01808v1.pdf

Episode sponsored by Tidelift: tidelift.comWe already mentioned multilevel regression and post-stratification (MRP, or Mister P) on this podcast, but we didn’t dedicate a full episode to explaining how it works, why it’s useful to deal with non-representative data, and what its limits are. Well, let’s do that now, shall we?To that end, I had the delight to talk with Lauren Kennedy! Lauren is a lecturer in Business Analytics at Monash University in Melbourne, Australia, where she develops new statistical methods to analyze social science data. Working mainly with R and Stan, Lauren studies non-representative data, multilevel modeling, post-stratification, causal inference, and, more generally, how to make inferences from the social sciences.Needless to say that I asked her everything I could about MRP, including how to choose priors, why her recent paper about structured priors can improve MRP, and when MRP is not useful. We also talked about missing data imputation, and how all these methods relate to causal inference in the social sciences.If you want a bit of background, Lauren did her Undergraduates in Psychological Sciences and Maths and Computer Sciences at Adelaide University, with Danielle Navarro and Andrew Perfors, and then did her PhD with the same advisors. She spent 3 years in NYC with Andrew Gelman’s Lab at Columbia University, and then moved back to Melbourne in 2020. Most importantly, Lauren is an adept of crochet — she’s already on her third blanket!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, Colin Carroll, Nathaniel Burbank, Michael Osthege and Rémi Louf.Visit https://www.patreon.com/learnbayesstats to unlock exclusive Bayesian swag ;)Links from the show:Lauren's website: https://jazzystats.com/Lauren on Twitter: https://twitter.com/jazzystatsLauren on GitHub: https://github.com/lauken13Improving multilevel regression and poststratification with structured priors: https://arxiv.org/abs/1908.06716Using model-based regression and poststratification to generalize findings beyond the observed sample: https://arxiv.org/abs/1906.11323Lauren's beginners Bayes workshop: https://github.com/lauken13/Beginners_Bayes_WorkshopMRP in RStanarm: https://github.com/lauken13/rstanarm/blob/master/vignettes/mrp.RmdChoosing your rstanarm prior with prior predictive checks: https://github.com/stan-dev/rstanarm/blob/vignette-prior-predictive/vignettes/prior-pred.RmdMister P -- What’s its secret sauce?: https://statmodeling.stat.columbia.edu/2013/10/09/mister-p-whats-its-secret-sauce/Bayesian Multilevel Estimation with Poststratification -- State-Level Estimates from National Polls: https://pdfs.semanticscholar.org/2008/bee9f8c2d7e41ac9c5c54489f41989a0d7ba.pdfMRPyMC3 - Multilevel Regression and Poststratification with PyMC3: https://austinrochford.com/posts/2017-07-09-mrpymc3.htmlUsing Hierarchical Multinomial Regression to Predict Elections in Paris districts: https://www.youtube.com/watch?v=EYdIzSYwbSwRegression and Other Stories book: https://www.cambridge.org/fr/academic/subjects/statistics-probability/statistical-theory-and-methods/regression-and-other-stories?format=PB Bayesian Nonparametric Modeling for Causal Inference, by Jennifer Hill: https://www.tandfonline.com/doi/abs/10.1198/jcgs.2010.08162Lauren's Data Ethics course: https://anastasiospanagiotelis.netlify.app/teaching/dataviza2019/lectures/04dataethics/ethicaldatascience#1