Today’s clip is from episode 137 of the podcast, with Robert Ness.

Alex and Robert discuss the intersection of causal inference and deep learning, emphasizing the importance of understanding causal concepts in statistical modeling.

The discussion also covers the evolution of probabilistic machine learning, the role of inductive biases, and the potential of large language models in causal analysis, highlighting their ability to translate natural language into formal causal queries.

Get the full conversation here.

Attend Alex's tutorial at PyData Berlin: A Beginner's Guide to State Space Modeling 

• Intro to Bayes Course (first 2 lessons free)
• Advanced Regression Course (first 2 lessons free)

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

Visit our Patreon page to unlock exclusive Bayesian swag ;)

Transcript

This is an automatic transcript and may therefore contain errors. Please get in touch if you're willing to correct them.

Proudly sponsored by PyMC Labs, the Bayesian Consultancy. Book a call, or get in touch!

• Intro to Bayes Course (first 2 lessons free)
• Advanced Regression Course (first 2 lessons free)

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

Visit our Patreon page to unlock exclusive Bayesian swag ;)

Takeaways:

• Causal assumptions are crucial for statistical modeling.
• Deep learning can be integrated with causal models.
• Statistical rigor is essential in evaluating LLMs.
• Causal representation learning is a growing field.
• Inductive biases in AI should match key mechanisms.
• Causal AI can improve decision-making processes.
• The future of AI lies in understanding causal relationships.

Chapters:

00:00 Introduction to Causal AI and Its Importance

16:34 The Journey to Writing Causal AI

28:05 Integrating Graphical Causality with Deep Learning

40:10 The Evolution of Probabilistic Machine Learning

44:34 Practical Applications of Causal AI with LLMs

49:48 Exploring Multimodal Models and Causality

56:15 Tools and Frameworks for Causal AI

01:03:19 Statistical Rigor in Evaluating LLMs

01:12:22 Causal Thinking in Real-World Deployments

01:19:52 Trade-offs in Generative Causal Models

01:25:14 Future of Causal Generative Modeling

Thank you to my Patrons for making this episode possible!

Yusuke Saito, Avi Bryant, Ero Carrera, Giuliano Cruz, James Wade, Tradd Salvo, William Benton, James Ahloy, Robin Taylor,, Chad Scherrer, 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, 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, Aubrey Clayton, Jeannine Sue, Omri Har Shemesh, Scott Anthony Robson, Robert Yolken, Or Duek, Pavel Dusek, Paul Cox, Andreas Kröpelin, Raphaël R, Nicolas Rode, Gabriel Stechschulte, Arkady, Kurt TeKolste, Marcus Nölke, Maggi Mackintosh, Grant Pezzolesi, Joshua Meehl, Javier Sabio, Kristian Higgins, Matt Rosinski, Bart Trudeau, Luis Fonseca, Dante Gates, Matt Niccolls, Maksim Kuznecov, Michael Thomas, Luke Gorrie, Cory Kiser, Julio, Edvin Saveljev, Frederick Ayala, Jeffrey Powell, Gal Kampel, Adan Romero, Will Geary, Blake Walters, Jonathan Morgan, Francesco Madrisotti, Ivy Huang, Gary Clarke, Robert Flannery, Rasmus Hindström, Stefan, Corey Abshire, Mike Loncaric, David McCormick, Ronald Legere, Sergio Dolia, Michael Cao, Yiğit Aşık, Suyog Chandramouli and Adam Tilmar Jakobsen.

Links from the show:

• Robert’s profile on Microsoft Research: https://www.microsoft.com/en-us/research/people/robertness/
• Robert’s Causal AI book: https://www.manning.com/books/causal-ai
• Associated GitHub repo: https://github.com/altdeep/causalML
• Robert on Google Scholar: https://scholar.google.com/citations?hl=en&user=8gWTOBAAAAAJ
• Robert on LinkedIn: https://www.linkedin.com/in/osazuwa/
• Robert on GitHub: https://github.com/robertness
• LBS #56 Causal & Probabilistic Machine Learning, with Robert Osazuwa Ness: https://learnbayesstats.com/episode/56-causal-probabilistic-machine-learning-robert-ness
• A Causal AI Suite for Decision-Making, NeurIPS 2022: https://openreview.net/forum?id=-gVJ1_lD1RH
• Causal Reasoning with ChiRho: https://basisresearch.github.io/chirho/getting_started.html
• PyWhy, An Open Source Ecosystem for Causal Machine Learning: https://www.pywhy.org/

Transcript

This is an automatic transcript and may therefore contain errors. Please get in touch if you're willing to correct them.

Today’s clip is from episode 136 of the podcast, with Haavard Rue & Janet van Niekerk.

Alex, Haavard and Janet explore the world of Bayesian inference with INLA, a fast and deterministic method that revolutionizes how we handle large datasets and complex models.

Discover the power of INLA, and why it can make your models go much faster! Get the full conversation here.

• Intro to Bayes Course (first 2 lessons free)
• Advanced Regression Course (first 2 lessons free)

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

Visit our Patreon page to unlock exclusive Bayesian swag ;)

Transcript

This is an automatic transcript and may therefore contain errors. Please get in touch if you're willing to correct them.

Proudly sponsored by PyMC Labs, the Bayesian Consultancy. Book a call, or get in touch!

• Intro to Bayes Course (first 2 lessons free)
• Advanced Regression Course (first 2 lessons free)

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

Visit our Patreon page to unlock exclusive Bayesian swag ;)

Takeaways:

• INLA is a fast, deterministic method for Bayesian inference.
• INLA is particularly useful for large datasets and complex models.
• The R INLA package is widely used for implementing INLA methodology.
• INLA has been applied in various fields, including epidemiology and air quality control.
• Computational challenges in INLA are minimal compared to MCMC methods.
• The Smart Gradient method enhances the efficiency of INLA.
• INLA can handle various likelihoods, not just Gaussian.
• SPDs allow for more efficient computations in spatial modeling.
• The new INLA methodology scales better for large datasets, especially in medical imaging.
• Priors in Bayesian models can significantly impact the results and should be chosen carefully.
• Penalized complexity priors (PC priors) help prevent overfitting in models.
• Understanding the underlying mathematics of priors is crucial for effective modeling.
• The integration of GPUs in computational methods is a key future direction for INLA.
• The development of new sparse solvers is essential for handling larger models efficiently.

Chapters:

06:06 Understanding INLA: A Comparison with MCMC

08:46 Applications of INLA in Real-World Scenarios

11:58 Latent Gaussian Models and Their Importance

15:12 Impactful Applications of INLA in Health and Environment

18:09 Computational Challenges and Solutions in INLA

21:06 Stochastic Partial Differential Equations in Spatial Modeling

23:55 Future Directions and Innovations in INLA

39:51 Exploring Stochastic Differential Equations

43:02 Advancements in INLA Methodology

50:40 Getting Started with INLA

56:25 Understanding Priors in Bayesian Models

Thank you to my Patrons for making this episode possible!

Yusuke Saito, Avi Bryant, Ero Carrera, Giuliano Cruz, James Wade, Tradd Salvo, William Benton, James Ahloy, Robin Taylor,, Chad Scherrer, 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, 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, Aubrey Clayton, Jeannine Sue, Omri Har Shemesh, Scott Anthony Robson, Robert Yolken, Or Duek, Pavel Dusek, Paul Cox, Andreas Kröpelin, Raphaël R, Nicolas Rode, Gabriel Stechschulte, Arkady, Kurt TeKolste, Marcus Nölke, Maggi Mackintosh, Grant Pezzolesi, Joshua Meehl, Javier Sabio, Kristian Higgins, Alex Jones, Matt Rosinski, Bart Trudeau, Luis Fonseca, Dante Gates, Matt Niccolls, Maksim Kuznecov, Michael Thomas, Luke Gorrie, Cory Kiser, Julio, Edvin Saveljev, Frederick Ayala, Jeffrey Powell, Gal Kampel, Adan Romero, Will Geary, Blake Walters, Jonathan Morgan, Francesco Madrisotti, Ivy Huang, Gary Clarke, Robert Flannery, Rasmus Hindström, Stefan, Corey Abshire, Mike Loncaric, David McCormick, Ronald Legere, Sergio Dolia, Michael Cao, Yiğit Aşık, Suyog Chandramouli and Adam Tilmar Jakobsen.

Links from the show:

• R-INLA webpage: https://www.r-inla.org/ 
• R-INLA discussion group: https://groups.google.com/g/r-inla-discussion-group 
• Haavard’s page: https://cemse.kaust.edu.sa/profiles/haavard-rue
• Haavard on Google Scholar: ​​https://scholar.google.co.uk/citations?user=VJOn_ZkAAAAJ&hl=en
• Janet’s page: https://cemse.kaust.edu.sa/profiles/janet-van-niekerk
• Janet on LinkedIn: https://www.linkedin.com/in/janet-van-niekerk-b1803b8a/
• Janet on Google Scholar: https://scholar.google.com/citations?user=rZOmGkAAAAAJ&hl=en
• Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations (original and classic formulation): https://users.wpi.edu/~balnan/INLAjrssB2009.pdf
• A new avenue for Bayesian inference with INLA (modern formulation of INLA and the current default in the R-INLA package): https://www.sciencedirect.com/science/article/pii/S0167947323000038 
• Smart Gradient - An adaptive technique for improving gradient estimation: https://www.aimsciences.org/article/doi/10.3934/fods.2021037 

SPDE-INLA book and other resources: 

• https://rss.onlinelibrary.wiley.com/doi/10.1111/j.1467-9868.2011.00777.x 
• https://becarioprecario.bitbucket.io/spde-gitbook/ 
• Inlabru: https://inlabru-org.github.io/inlabru/ 

Penalizing complexity priors: 

• Penalizing Model Component Complexity: A Principled, Practical Approach to Constructing Priors: https://doi.org/10.1214/16-STS576
• AR processes  https://doi.org/10.1111/jtsa.12242 
• Gaussian fields https://doi.org/10.1080/01621459.2017.1415907
• Skew-normal model https://doi.org/10.57805/revstat.v19i1.328
• Weibull model https://doi.org/10.1016/j.spl.2021.109098
• Splines https://arxiv.org/abs/1511.05748
• Many more available in the R-INLA library: inla.pc

Transcript

This is an automatic transcript and may therefore contain errors. Please get in touch if you're willing to correct them.

Get 10% off Hugo's "Building LLM Applications for Data Scientists and Software Engineers" online course!

Today’s clip is from episode 135 of the podcast, with Teemu Säilynoja.

Alex and Teemu discuss the importance of simulation-based calibration (SBC). They explore the practical implementation of SBC in probabilistic programming languages, the challenges faced in developing SBC methods, and the significance of both prior and posterior SBC in ensuring model reliability.

The discussion emphasizes the need for careful model implementation and inference algorithms to achieve accurate calibration.

Get the full conversation here.

• Intro to Bayes Course (first 2 lessons free)
• Advanced Regression Course (first 2 lessons free)

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

Visit our Patreon page to unlock exclusive Bayesian swag ;)

Transcript

This is an automatic transcript and may therefore contain errors. Please get in touch if you're willing to correct them.

Proudly sponsored by PyMC Labs, the Bayesian Consultancy. Book a call, or get in touch!

• Intro to Bayes Course (first 2 lessons free)
• Advanced Regression Course (first 2 lessons free)

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

Visit our Patreon page to unlock exclusive Bayesian swag ;)

Takeaways:

• Teemu focuses on calibration assessments and predictive checking in Bayesian workflows.
• Simulation-based calibration (SBC) checks model implementation
• SBC involves drawing realizations from prior and generating prior predictive data.
• Visual predictive checking is crucial for assessing model predictions.
• Prior predictive checks should be done before looking at data.
• Posterior SBC focuses on the area of parameter space most relevant to the data.
• Challenges in SBC include inference time.
• Visualizations complement numerical metrics in Bayesian modeling.
• Amortized Bayesian inference benefits from SBC for quick posterior checks. The calibration of Bayesian models is more intuitive than Frequentist models.
• Choosing the right visualization depends on data characteristics.
• Using multiple visualization methods can reveal different insights.
• Visualizations should be viewed as models of the data.
• Goodness of fit tests can enhance visualization accuracy.
• Uncertainty visualization is crucial but often overlooked.

Chapters:

09:53 Understanding Simulation-Based Calibration (SBC)

15:03 Practical Applications of SBC in Bayesian Modeling

22:19 Challenges in Developing Posterior SBC

29:41 The Role of SBC in Amortized Bayesian Inference

33:47 The Importance of Visual Predictive Checking

36:50 Predictive Checking and Model Fitting

38:08 The Importance of Visual Checks

40:54 Choosing Visualization Types

49:06 Visualizations as Models

55:02 Uncertainty Visualization in Bayesian Modeling

01:00:05 Future Trends in Probabilistic Modeling

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, 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, 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, Andreas Kröpelin, Raphaël R, Nicolas Rode, Gabriel Stechschulte, Arkady, Kurt TeKolste, Gergely Juhasz, Marcus Nölke, Maggi Mackintosh, Grant Pezzolesi, Avram Aelony, Joshua Meehl, Javier Sabio, Kristian Higgins, Alex Jones, Gregorio Aguilar, Matt Rosinski, Bart Trudeau, Luis Fonseca, Dante Gates, Matt Niccolls, Maksim Kuznecov, Michael Thomas, Luke Gorrie, Cory Kiser, Julio, Edvin Saveljev, Frederick Ayala, Jeffrey Powell, Gal Kampel, Adan Romero, Will Geary, Blake Walters, Jonathan Morgan, Francesco Madrisotti, Ivy Huang, Gary Clarke, Robert Flannery, Rasmus Hindström, Stefan, Corey Abshire, Mike Loncaric, David McCormick, Ronald Legere, Sergio Dolia, Michael Cao, Yiğit Aşık and Suyog Chandramouli.

Links from the show:

• Teemu's website: https://teemusailynoja.github.io/
• Teemu on LinkedIn: https://www.linkedin.com/in/teemu-sailynoja/
• Teemu on GitHub: https://github.com/TeemuSailynoja
• Bayesian Workflow group: https://users.aalto.fi/~ave/group.html
• LBS #107 Amortized Bayesian Inference with Deep Neural Networks, with Marvin Schmitt: https://learnbayesstats.com/episode/107-amortized-bayesian-inference-deep-neural-networks-marvin-schmitt
• LBS #73 A Guide to Plotting Inferences & Uncertainties of Bayesian Models, with Jessica Hullman: https://learnbayesstats.com/episode/73-guide-plotting-inferences-uncertainties-bayesian-models-jessica-hullman
• LBS #66 Uncertainty Visualization & Usable Stats, with Matthew Kay: https://learnbayesstats.com/episode/66-uncertainty-visualization-usable-stats-matthew-kay
• LBS #35 The Past, Present & Future of BRMS, with Paul Bürkner: https://learnbayesstats.com/episode/35-past-present-future-brms-paul-burkner
• LBS #29 Model Assessment, Non-Parametric Models, And Much More, with Aki Vehtari: https://learnbayesstats.com/episode/model-assessment-non-parametric-models-aki-vehtari
• Posterior SBC – Simulation-Based Calibration Checking Conditional on Data: https://arxiv.org/abs/2502.03279
• Recommendations for visual predictive checks in Bayesian workflow: https://teemusailynoja.github.io/visual-predictive-checks/
• Simuk, SBC for PyMC: https://simuk.readthedocs.io/en/latest/
• SBC, tools for model validation in R: https://hyunjimoon.github.io/SBC/index.html 
• New ArviZ, Prior and Posterior predictive checks: https://arviz-devs.github.io/EABM/Chapters/Prior_posterior_predictive_checks.html
• Bayesplot, plotting for Bayesian models in R: https://mc-stan.org/bayesplot/

Transcript

This is an automatic transcript and may therefore contain errors. Please get in touch if you're willing to correct them.

ICYMI, I'll be in London next week, for a live episode of the Learning Bayesian Statistics podcast 🍾

Come say hi on June 24 at Imperial College London! We'll be talking about uncertainty quantification — not just in theory, but in the messy, practical reality of building models that are supposed to work in the real world.

🎟️ Get your tickets!

Some of the questions we’ll unpack:

🔍 Why is it so hard to model uncertainty reliably?

⚠️ How do overconfident models break things in production?

🧠 What tools and frameworks help today?

🔄 What do we need to rethink if we want robust ML over the next decade?

Joining me on stage: the brilliant Mélodie Monod, Yingzhen Li and François-Xavier Briol -- researchers doing cutting-edge work on these questions, across Bayesian methods, statistical learning, and real-world ML deployment.

A huge thank you to Oliver Ratmann for setting this up!

📍 Imperial-X, White City Campus (Room LRT 608)

🗓️ June 24, 11:30–13:00

🎙️ Doors open at 11:30 — we start at noon sharp

Come say hi, ask hard questions, and be part of the recording.

🎟️ Get your tickets!

• Intro to Bayes Course (first 2 lessons free)
• Advanced Regression Course (first 2 lessons free)

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

Visit our Patreon page to unlock exclusive Bayesian swag ;)

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, 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, 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, Andreas Kröpelin, Raphaël R, Nicolas Rode, Gabriel Stechschulte, Arkady, Kurt TeKolste, Gergely Juhasz, Marcus Nölke, Maggi Mackintosh, Grant Pezzolesi, Avram Aelony, Joshua Meehl, Javier Sabio, Kristian Higgins, Alex Jones, Gregorio Aguilar, Matt Rosinski, Bart Trudeau, Luis Fonseca, Dante Gates, Matt Niccolls, Maksim Kuznecov, Michael Thomas, Luke Gorrie, Cory Kiser, Julio, Edvin Saveljev, Frederick Ayala, Jeffrey Powell, Gal Kampel, Adan Romero, Will Geary, Blake Walters, Jonathan Morgan, Francesco Madrisotti, Ivy Huang, Gary Clarke, Robert Flannery, Rasmus Hindström, Stefan, Corey Abshire, Mike Loncaric, David McCormick, Ronald Legere, Sergio Dolia, Michael Cao, Yiğit Aşık and Suyog Chandramouli.

Transcript

This is an automatic transcript and may therefore contain errors. Please get in touch if you're willing to correct them.

Today’s clip is from episode 134 of the podcast, with David Kohns.

Alex and David discuss the future of probabilistic programming, focusing on advancements in time series modeling, model selection, and the integration of AI in prior elicitation.

The discussion highlights the importance of setting appropriate priors, the challenges of computational workflows, and the potential of normalizing flows to enhance Bayesian inference.

Get the full discussion here.

• Intro to Bayes Course (first 2 lessons free)
• Advanced Regression Course (first 2 lessons free)

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

Visit our Patreon page to unlock exclusive Bayesian swag ;)

Transcript

This is an automatic transcript and may therefore contain errors. Please get in touch if you're willing to correct them.

Proudly sponsored by PyMC Labs, the Bayesian Consultancy. Book a call, or get in touch!

• Intro to Bayes Course (first 2 lessons free)
• Advanced Regression Course (first 2 lessons free)

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

Visit our Patreon page to unlock exclusive Bayesian swag ;)

Takeaways:

• Setting appropriate priors is crucial to avoid overfitting in models.
• R-squared can be used effectively in Bayesian frameworks for model evaluation.
• Dynamic regression can incorporate time-varying coefficients to capture changing relationships.
• Predictively consistent priors enhance model interpretability and performance.
• Identifiability is a challenge in time series models.
• State space models provide structure compared to Gaussian processes.
• Priors influence the model's ability to explain variance.
• Starting with simple models can reveal interesting dynamics.
• Understanding the relationship between states and variance is key.
• State-space models allow for dynamic analysis of time series data.
• AI can enhance the process of prior elicitation in statistical models.

Chapters:

10:09 Understanding State Space Models

14:53 Predictively Consistent Priors

20:02 Dynamic Regression and AR Models

25:08 Inflation Forecasting

50:49 Understanding Time Series Data and Economic Analysis

57:04 Exploring Dynamic Regression Models

01:05:52 The Role of Priors

01:15:36 Future Trends in Probabilistic Programming

01:20:05 Innovations in Bayesian Model Selection

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, 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, 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, Andreas Kröpelin, Raphaël R, Nicolas Rode, Gabriel Stechschulte, Arkady, Kurt TeKolste, Gergely Juhasz, Marcus Nölke, Maggi Mackintosh, Grant Pezzolesi, Avram Aelony, Joshua Meehl, Javier Sabio, Kristian Higgins, Alex Jones, Gregorio Aguilar, Matt Rosinski, Bart Trudeau, Luis Fonseca, Dante Gates, Matt Niccolls, Maksim Kuznecov, Michael Thomas, Luke Gorrie, Cory Kiser, Julio, Edvin Saveljev, Frederick Ayala, Jeffrey Powell, Gal Kampel, Adan Romero, Will Geary, Blake Walters, Jonathan Morgan, Francesco Madrisotti, Ivy Huang, Gary Clarke, Robert Flannery, Rasmus Hindström, Stefan, Corey Abshire, Mike Loncaric, David McCormick, Ronald Legere, Sergio Dolia, Michael Cao, Yiğit Aşık and Suyog Chandramouli.

Links from the show:

• David's website: https://davkoh.github.io/
• David on LinkedIn: https://www.linkedin.com/in/david-kohns-03984013b/
• David on GitHub: https://github.com/davkoh
• David on Google Scholar: https://scholar.google.com/citations?user=9gKE8e4AAAAJ&hl=en
• Dynamic Regression Case Study: https://davkoh.github.io/case-studies/01_dyn_reg/dyn_reg_casestudy5.html
• ARR2 Paper: https://projecteuclid.org/journals/bayesian-analysis/advance-publication/The-ARR2-Prior--Flexible-Predictive-Prior-Definition-for-Bayesian/10.1214/25-BA1512.full
• ARR2 Paper GitHub repository: https://github.com/n-kall/arr2/tree/main
• ARR2 StanCon talk: https://www.youtube.com/watch?v=8XBe2jrOKvw&list=PLCrWEzJgSUqzNzh6mjWsWUu-lSK59VXP6&index=29
• ARR2 Prior in PyMC: https://www.austinrochford.com/posts/r2-priors-pymc.html
• LBS #124 State Space Models & Structural Time Series, with Jesse Grabowski: https://learnbayesstats.com/episode/124-state-space-models-structural-time-series-jesse-grabowski
• LBS #107 Amortized Bayesian Inference with Deep Neural Networks, with Marvin Schmitt: https://learnbayesstats.com/episode/107-amortized-bayesian-inference-deep-neural-networks-marvin-schmitt
• LBS #74 Optimizing NUTS and Developing the ZeroSumNormal Distribution, with Adrian Seyboldt: https://learnbayesstats.com/episode/74-optimizing-nuts-developing-zerosumnormal-distribution-adrian-seyboldt
• Nutpie’s Normalizing Flows adaptation: https://pymc-devs.github.io/nutpie/nf-adapt.html

Transcript

This is an automatic transcript and may therefore contain errors. Please get in touch if you're willing to correct them.

Today’s clip is from episode 133 of the podcast, with Sean Pinkney & Adrian Seyboldt.

The conversation delves into the concept of Zero-Sum Normal and its application in statistical modeling, particularly in hierarchical models.

Alex, Sean and Adrian discuss the implications of using zero-sum constraints, the challenges of incorporating new data points, and the importance of distinguishing between sample and population effects.

They also explore practical solutions for making predictions based on population parameters and the potential for developing tools to facilitate these processes.

Get the full discussion here.

• Intro to Bayes Course (first 2 lessons free)
• Advanced Regression Course (first 2 lessons free)

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

Visit our Patreon page to unlock exclusive Bayesian swag ;)

Transcript

This is an automatic transcript and may therefore contain errors. Please get in touch if you're willing to correct them.