Lightnews — Scholar-powered news

PyMC Labs @pymc-labs.bsky.social · 2d

Can #LLMs really estimate prices, and why does it matter for business?

At #PyMC Labs, we built a “Price is Right” inspired benchmark testing models’ price knowledge & strategy.

Results show accuracy isn’t enough; reasoning matters.

Explore 👉 dub.link/CbLASs6

1

PyMC Labs @pymc-labs.bsky.social · 6d

You’ve been saying you’ll learn #Bayesian modeling someday. Make someday 𝐭𝐨𝐝𝐚𝐲.

The next 𝐀𝐩𝐩𝐥𝐢𝐞𝐝 𝐁𝐚𝐲𝐞𝐬𝐢𝐚𝐧 𝐌𝐨𝐝𝐞𝐥𝐢𝐧𝐠 𝐖𝐨𝐫𝐤𝐬𝐡𝐨𝐩 from #PyMC Labs begins October 6 and registration closes this week.

Secure your spot now: dub.link/yhoQ7tF

1 1

PyMC Labs @pymc-labs.bsky.social · 7d

Starting in just 10 minutes: 𝐏𝐲𝐌𝐂-𝐌𝐚𝐫𝐤𝐞𝐭𝐢𝐧𝐠 𝐯𝐬. 𝐆𝐨𝐨𝐠𝐥𝐞 𝐌𝐞𝐫𝐢𝐝𝐢𝐚𝐧 – A Scientific Benchmark for MMM.

Don’t miss this session where our team shares the results of a rigorous, side-by-side comparison of two leading MMM libraries, helping you make evidence-based decisions for your modeling goals.

PyMC Labs @pymc-labs.bsky.social · 7d

Final session of our #PyTensor Workshop is live!

Discover custom rewrites, speed up diagonal matrix inversions 80x, and see a mini-PyMC in action - one graph powering the full Bayesian workflow.

Watch here: dub.link/WRFuYtp

#PyMC #Python #ML #DataScience

PyTensor Workshop #3 | Custom Rewrites & Building a Mini-PyMC

Enjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.

dub.link

1 3

PyMC Labs @pymc-labs.bsky.social · 14d

Join Rob Zinkov at #PyDataAmsterdam 10th Anniversary on Sep 26, 1:20 PM for a beginner-friendly #PyMC sprint.

Learn, contribute, and connect with the community!

#PyTensor

2

PyMC Labs @pymc-labs.bsky.social · 15d

Last week, we published the most comprehensive #PyMC-Marketing vs. #Meridian benchmark to date.

𝐎𝐧 𝐎𝐜𝐭 𝟏, join our 𝐥𝐢𝐯𝐞 𝐰𝐞𝐛𝐢𝐧𝐚𝐫 to go deeper into the findings & their impact on #MMM.

🗓️ Register: dub.link/rgm4PzI
👉Check the Benchmark: dub.link/CLWnkdf

1 1 1

PyMC Labs @pymc-labs.bsky.social · 16d

When learning #Bayesian modeling, great teaching 𝑎𝑛𝑑 great resources matter.

Hear from Bartłomiej, a workshop alum, on how the Applied Bayesian Modeling Workshop gave him practical tools to keep using.

𝐍𝐞𝐱𝐭 𝐰𝐨𝐫𝐤𝐬𝐡𝐨𝐩: 𝐎𝐜𝐭 𝟔 → dub.link/UuDaiQk

1

PyMC Labs @pymc-labs.bsky.social · 20d

Most ML models predict, but don’t show how certain those predictions are.

#Bayesian methods with #PyMC change that.

Tonight in Montreal, Christopher Fonnesbeck will show how to model uncertainty with practical workflows + live coding.

🔗 dub.link/P2iaD9M

1 3

PyMC Labs @pymc-labs.bsky.social · 21d

Session 2 of our #PyTensor workshop dives into 𝐆𝐫𝐚𝐩𝐡 𝐓𝐫𝐚𝐧𝐬𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧𝐬!

Learn eager vs. lazy transformations, autodiff under the hood, and Normalizing Flows.

Watch the full session here: dub.link/Ihd3Zc6

#PyMC #ML

PyTensor Workshop#2 | Advanced Graph Transformations & Normalizing Flows

Enjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.

dub.link

3

PyMC Labs @pymc-labs.bsky.social · 22d

𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 𝐁𝐚𝐲𝐞𝐬𝐢𝐚𝐧 𝐌𝐨𝐝𝐞𝐥𝐢𝐧𝐠 𝐨𝐧 𝐲𝐨𝐮𝐫 𝐨𝐰𝐧 𝐜𝐚𝐧 𝐛𝐞 𝐬𝐥𝐨𝐰.

Hear how Leonardo leveled up in the Applied Bayesian Modeling Workshop.

Next cohort starting Oct 6: dub.link/admGEW9

#BayesianModeling #PyMC #MCMC

1

PyMC Labs @pymc-labs.bsky.social · 27d

We just dropped the most comprehensive #PyMC-Marketing vs. #Meridian 𝐐𝐮𝐚𝐧𝐭𝐢𝐭𝐚𝐭𝐢𝐯𝐞 𝐂𝐨𝐦𝐩𝐚𝐫𝐢𝐬𝐨𝐧 ever conducted!

Check out our findings & code 👇👇
dub.link/CLWnkdf

#MMM #MarketingMixModel #PyMCMarketing #Meridian

PyMC-Marketing vs. Meridian: A Quantitative Comparison of Open Source MMM Libraries

This benchmark study directly compares PyMC-Marketing and Google’s Meridian on realistic synthetic datasets, from startup-scale to enterprise-level. Using aligned priors and identical configurations, ...

dub.link

1

PyMC Labs @pymc-labs.bsky.social · 28d

Behind every #PyMC model is #PyTensor — a computational graph library for speed, flexibility & autodiff.

Session 1 of our PyTensor Workshop is live 🎥
👉 dub.link/pytensor-gra...

#PyMC #Python #MachineLearning

- YouTube

Enjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.

dub.link

3

PyMC Labs @pymc-labs.bsky.social · 29d

Most data scientists model. Few master uncertainty.

Join our Oct 6 Applied #Bayesian Modeling Workshop & level up.

👉 dub.link/QWC7F2N

1

PyMC Labs @pymc-labs.bsky.social · Sep 8

PyMC-Marketing 0.16.0 is live!
Sensitivity analysis, #Fivetran loaders & more.

Available now on GitHub
👉 dub.link/bGBLG2V

#Bayesian #MMM #PyMC

PyMC Labs @pymc-labs.bsky.social · Sep 5

We had an incredible time at PyData Berlin ✨

Such a vibrant community, from the conversations and presentations to the speakers and attendees, the energy was just amazing.

A huge thank you to @pydata.bsky.social for continuing to nurture a community that inspires and connects us all!🙌

1 3

PyMC Labs @pymc-labs.bsky.social · Sep 4

Stop waiting weeks for modeling answers!

Our new Expert Access Program (EAP) gives your team direct access to our #Bayesian and #PyMC experts for timely answers, guidance, and confidence to move faster.

Learn more👉 dub.link/qKrWGwa

PyMC Labs @pymc-labs.bsky.social · Sep 3

🤝 Announcing PyMC Labs x @fivetran.bsky.social Partnership: Marketing Mix Models in minutes, not weeks

Data prep was the biggest #MMM bottleneck. We solved it.

✓ Direct integration with Fivetran ad_reporting
✓ Production-grade Bayesian MMMs instantly
👉 dub.link/csv37dq

1

PyMC Labs @pymc-labs.bsky.social · Sep 2

Our Applied #Bayesian Modeling Workshop - October cohort is now open!

After incredible feedback from August, we're running it again starting Oct 6th. Learn directly from the creators of #PyMC in 4 weeks of hands-on training!

Register before spots fill: dub.link/QWC7F2N

4

PyMC Labs @pymc-labs.bsky.social · Aug 28

Enlightened monks. Crying babies. Data scientists

👀 All bound by the same trauma: p-values

The too-high one
The model that made no sense
The sketchy dataset you used anyway
Bayes won’t absolve you—but it will help you understand

👇 Step into the Bayesian Booth. Confess.

#BayesianMethods #pvalues

PyMC Labs @pymc-labs.bsky.social · Aug 27

Time series models work beautifully on textbook data. Real data? Not so much.

State Space Models are built for the messiness of production data from day one.

Learn to build them with Alexandre Andorra and Jesse Grabowski at #PyData Berlin 2025

📈 Sept 1st | Room B09 | 10:40 AM

3

PyMC Labs @pymc-labs.bsky.social · Aug 26

For years, we kept hearing the same request: "Can we just call you when we get stuck?" So we made it official.

Announcing the launch of our Expert Access Program

Get direct access to #PyMC experts for guidance, validation, & problem-solving without adding headcount!

dub.link/BMHefW7

Announcing the Expert Access Program (EAP)

Teams face critical deadlines where models must be rigorous, but hiring consultants or adding headcount only solves problems temporarily. This is why we created the Expert Access Program — to give tea...

dub.link

1 2

PyMC Labs @pymc-labs.bsky.social · Aug 25

Market share predictions fail when models ignore how markets really work.

At #PyDataBerlin, @nathanielforde.bsky.social shows how PyMC Marketing’s new Consumer Choice module uses Bayesian models for realistic substitution.

🗓️ Sept 1, 4:20 PM | 📍 Room B05-B06
#BayesianStats #PyMC

1 2

PyMC Labs @pymc-labs.bsky.social · Aug 21

Most budget planning is just expensive guessing 🎯

Carlos Eduardo Trujillo Agostini tackles this in "Risk Budget Optimization for Causal Mix Models" @PyDataBerlin 2025!

📅 Sept 1st, 5:00-5:30 PM

Turn uncertainty into strategic advantage.

#PyDataBerlin #BudgetOptimization #PyMC

1 2

PyMC Labs @pymc-labs.bsky.social · Aug 19

When your #Bayesian model hits production and #MCMC crawls for hours, the bottleneck isn't your modeling skills; it's your inference engine.

@juanitorduz.bsky.social will tackle this in "Scaling Probabilistic Models with Variational Inference" PyData Berlin 📅 Sept 2nd, 1:40 PM

2 8

PyMC Labs @pymc-labs.bsky.social · Aug 18

We're heading to #PyDataBerlin 2025 (Sept 1-3)!

Our team will be presenting on state space models, variational inference, and probabilistic approaches to marketing and risk.

Come find us if you're attending, we're always up for great conversations about open source data tools!
#DataScience #PyMC

3