Pierre-Simon Laplace

Models need more than pattern-matching.
They need causal understanding.

In this episode, Robert Ness joins @alex-andorra.bsky.social to explore:

⚡ Why models need real-world biases
🧠 How causal rep learning is reshaping AI
🤖 What it takes to add causality to DL

🎧 lnkd.in/gUnCkwEP

#bayes #podcast

🚨 MCMC or INLA?

🤯 MCMC = slow sampling.
⚡ INLA = fast, smart approximations. No chains, no waiting.

🎙️ On LBS, @alex-andorra.bsky.social talks with Haavard Rue & Janet Van Niekerk about how INLA works, when to use it, and why it’s a game-changer.

🎧 Listen: lnkd.in/gp8D-RuU

#Bayesian #MCMC

🚨 Tired of MCMC cooking your CPU for hours?

@alex-andorra.bsky.social chats with Haavard Rue & Janet van Niekerk about INLA, a fast, deterministic game-changer for inference at scale.

✅ Handles huge + complex models
✅ Works with non-Gaussian likelihoods

🎧 www.learnbayesstats.com/episode/136-...

🧲 Got 50 predictors, but only 5 that matter?

Try the Horseshoe Prior — a Bayesian approach to sparse regression that shrinks noise, not signal.

Built with Bambi + @pymc.io

🔗 Full demo: bambinos.github.io/bambi/notebo...

#BayesianStatistics #Regression #HorseshoePrior #MarketingAnalytics #PyMC

New episode is out! A very practical one, where we dive into *how* to make sure your models *actually* answer the questions you're asking...

🔍 Most Bayesian models aren’t properly checked

Even when they converge, they might be wrong in ways you won’t see—unless you look differently

In this episode, Teemu Säilynoja joins @alex-andorra.bsky.social to explore, SBC, prior predictive checks and more!

🎧 learnbayesstats.com/episode/135-...

Your model says 97% confidence
But should you trust it?

Uncertainty in ML is still a hard problem

We’re hosting a meetup at Imperial College London on June 24 to dig into it — with our host @alex-andorra.bsky.social and other researchers working on better ways forward

🔗 lnkd.in/eainEJ9p

New episode is out! In this one we nerd out quite deep on zero-sum constraints, and how to make your model sample faster 💨

Thanks again for the guest star appearance @aseyboldt.bsky.social !! You're welcome back anytime ;)

Some people think R² doesn’t belong in Bayesian models
👇 David Kohns disagrees, and he has the math to back it

🎙️Ep. 134: @alex-andorra.bsky.social sits down with economist David Kohns to explore how modern Bayesian methods are reshaping time series modelling

🎧 learnbayesstats.com/episode/134-...

Ask me any questions you may have about this! #stats

Learning Bayesian Statistics – Laplace to be for new & veteran Bayesians alike! Laplace to be for new & veteran Bayesians alike!

🎙️ Ep. 133 is out now!

@alex-andorra.bsky.social chats with ‪ @spinkney.bsky.social
& Adrian Seyboldt about making Bayesian models more efficient without losing rigor — zero-sum constraints, Cholesky tricks, practical wins & more

🎧 learnbayesstats.com/episode/133-...

#Bayesianstats #podcast #LBS

🎙️ In episode #132 of LBS, @alex-andorra.bsky.social talks with Tom Griffiths about Bayesian cognition and human-AI interaction—how we learn from limited data, why priors matter, what AI gets wrong, and why solving real problems beats perfect models & more ...

🔗 learnbayesstats.com/episode/132-...

⚽️ New Learning Bayesian Stats ep!

@alex-andorra.bsky.social & Luke Bornn dive into how tracking data, probabilistic models & optimization are reshaping sports decisions.

🎧 Listen now: learnbayesstats.com/episode/131-...

#Datascience #Optimization #SportsAnalytics #BayesStats
#Decisionmaking

The ARR2 Prior: Flexible Predictive Prior Definition for Bayesian Auto-Regressions We present the ARR2 prior, a joint prior over the auto-regressive components in Bayesian time-series models and their induced R2. Compared to other priors designed for times-series models, the ARR2 prior allows for flexible and intuitive shrinkage. We derive the prior for pure auto-regressive models, and extend it to auto-regressive models with exogenous covariates, and state-space models. Through both simulations and real-world modelling exercises, we demonstrate the efficacy of the ARR2 prior in improving sparse and reliable inference, while showing greater inference quality and predictive performance than other shrinkage priors. An open-source implementation of the prior is provided.

David will demo #TimeSeries models -- AR, MA, ARMA, auto-regressive distributed lag (ARDL) and vector auto-regressive (VAR) models.

👉 Paper: projecteuclid.org/journals/bay...

🚨 LIVE SHOW ALERT 🚨
If you're a Patron, join us tomorrow April 24, 11:00am, US Eastern Time, with David Kohns, co-author of "The ARR2 Prior: Flexible Predictive Prior Definition for Bayesian Auto-Regressions" 🥳
👉 Patreon: www.patreon.com/c/learnbayes...
👉 YouTube: www.youtube.com/@learningbay...

Enjoyed this wide ranging discussion of how we use data and models in epidemic response:

“I think of Bayesian Deep Learning as Bayesian inference in the function space”

This 55 minute interview is a really good primer of the current research questions in this field, definitely recommend it!

Thanks Claire, happy you liked it!

Hope you liked it!

🧬 What does real-world impact look like when public health’s on the line?

🎙️ In episode 130 of LBS, @alex-andorra.bsky.social chats with Adam Kucharski on modelling, crisis response & lessons from recent epidemics.

🎧 Listen in: learnbayesstats.com/episode/130-...

Couldn't be more thrilled to have you both among my audience 🫶

With his research on Bayesian Deep Learning, Vincent has his finger on the pulse of the time in modern AI 🤖

I’m already looking forward to tune in to this conversation between @alex-andorra.bsky.social and @vincefort.bsky.social.

… and it also comes with a video version! 👇

youtu.be/3hYYGiucS0U

This was a really fun chat with @alex-andorra.bsky.social, thanks for having me on the podcast! 🙂

What if AI could know when it doesn’t know?

🎙️ @alex-andorra.bsky.social talks with @vincefort.bsky.social about Bayesian deep learning, why it matters for uncertainty, calibration & more

🎧 Tune in: learnbayesstats.com/episode/129-...

#BayesianDeepLearning #MachineLearning #ReliableAI #AIResearch

⚽ Football is more than tactics and talent—it's driven by 𝐝𝐚𝐭𝐚. From 𝐩𝐥𝐚𝐲𝐞𝐫 𝐫𝐞𝐜𝐫𝐮𝐢𝐭𝐦𝐞𝐧𝐭 to 𝐦𝐚𝐭𝐜𝐡 𝐚𝐧𝐚𝐥𝐲𝐬𝐢𝐬, data science gives clubs a winning edge.

🎧 In the latest episode, @alex-andorra.bsky.social sits down with Matthew Penn to break it all down:

👉 learnbayesstats.com/episode/128-...

Thanks Lucas! One of our favorites too ;)

🌊 𝐒𝐡𝐚𝐫𝐤𝐬, 𝐏𝐲𝐭𝐡𝐨𝐧, 𝐚𝐧𝐝 𝐂𝐚𝐮𝐬𝐚𝐥 𝐈𝐧𝐟𝐞𝐫𝐞𝐧𝐜𝐞? 𝐘𝐞𝐬, 𝐲𝐨𝐮 𝐫𝐞𝐚𝐝 𝐭𝐡𝐚𝐭 𝐫𝐢𝐠𝐡𝐭!

🎧 In this episode, with Aaron MacNeil and @alex-andorra.bsky.social we dive deep (pun intended) into how advanced statistical methods play a surprising role in ocean conservation

🔗 learnbayesstats.com/episode/127-...

#Ecology

Field of Play UK | Sports data analytics conference Ready to level up your sports analytics game? Attend our sports data conference on 18th March run by Field of Play UK

⚾ @fonnesbeck.bsky.social (@pymc-labs.bsky.social, @pymc.io) will be at the Field of Play Conference giving a talk on Bayesian modelling in baseball.

😎 Our host, @alex-andorra.bsky.social , will also be attending, don’t miss this chance to connect and chat research!

🔗 www.fieldofplay.co.uk

📢 Episode 126 is Live!

🎧 Listen now 👉 learnbayesstats.com/episode/126-...

🎙️ In this episode with @alex-andorra.bsky.social , Will Dean from @pymc-labs.bsky.social explains how Bayesian methods are reshaping marketing analytics, from MMM to CLV estimation and more ....